of Sussex PhD thesissro.sussex.ac.uk/67293/1/Macklyne, Heather-Rose Victoria.pdf · enable strain development for second generation biofuel production. We have focused on the redox-sensing

A University of Sussex PhD thesis

Available online via Sussex Research Online:

http://sro.sussex.ac.uk/

This thesis is protected by copyright which belongs to the author.

This thesis cannot be reproduced or quoted extensively from without first obtaining permission in writing from the Author

The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the Author

When referring to this work, full bibliographic details including the author, title, awarding institution and date of the thesis must be given

Please visit Sussex Research Online for more information and further details

Engineering Bacteria for Biofuel Production

Heather-Rose Victoria Macklyne

Thesis submitted for the degree of

Doctor of Philosophy

University of Sussex

September 2016

3

Summary

This thesis addresses the need for environmentally and socially responsible sources

of energy. Biofuels, made from organic matter, have recently become a viable

alternative to petroleum-based fossil fuel. Sugar and starch make up the majority of

feedstock used in biofuel production as it is easily digested. However, the use of

these feedstocks is problematic as they consume resources with negative

implications. By using a bacterium able to utilise five and six carbon sugars, such as

the thermophile Geobacillus thermoglucosidans, organic lignocellulosic waste material

can be used as a feedstock.

The aim of this project was to investigate and utilise key genetic regulators of

fermentation in G. thermoglucosidans and to construct genetic engineering tools that

enable strain development for second generation biofuel production. We have

focused on the redox-sensing transcriptional regulator Rex, widespread in Gram-

positive bacteria, which controls the major fermentation pathways in response to

changes in cellular NAD+/NADH ratio. Following the identification of several

members of the Rex regulon via bioinformatics analysis, ChIP-seq and qRT-PCR

experiments were performed to locate genome-wide binding sites and controlled

genes in G. thermoglucosidans. Initial electromobility shift assay experiments were

performed to demonstrate the potential for use of Rex from Clostridium thermocellum

as an orthogonal regulator. To further this research, novel in vivo synthetic

regulatory switches were designed and tested with the aim of controlling gene

expression in response to changes in cellular redox state. In addition, new tools for

the efficient genetic engineering of G. thermoglucosidans were produced and

optimised, including an E. coli-G. thermoglucosidans conjugation method for plasmid

transfer and gene disruption.

4

Acknowledgments

Firstly I would like to thank Dr Mark Paget for the opportunity to work in his lab,

and acknowledge the huge amount of time and energy he has dedicated to helping

me during my PhD. To all members of the Paget research group, past and present,

including Heena Jagatia, Laurence Humphrey, Aline Tabib-Salazar: thank you for

your daily support and encouragement. I would also like to extend my thanks to the

Morley research group, including Ella Lineham, Alice Copsey and Simon Cpoper

for your friendship. I wish to thank my co-supervisor Dr Neil Crickmore for his help

and guidance throughout this project. I will think back on my time at Sussex fondly

and miss all the friends, fellow students and staff members who made it such an

enjoyable place to work and study.

The Student Support team has been an incredibly important part of my time at

Sussex and I want to acknowledge that their work has enabled me to succeed. I am

particularly grateful to my dyslexia tutors Moira Wilson and Alison Firbank for

believing in me, even when I didn’t believe in myself. I could never have achieved

this without you.

For the funding of this research and my PhD studentship I am thankful to the

BBSRC. I am also thankful for initial sponsorship and support from TMO

Renewables Ltd, particularly to Dr Alex Pudney, Gareth Cooper, Kirstin Eley. I

would also like to acknowledge Joanne Neary, Kevin Charman and Zaneta

Walanus for their help whilst visiting Rebio Technologies Ltd.

Ursula Gooding deserves my gratitude for making sure I didn’t fall through the

cracks when I left home at a young age; her support was vital and always offered

graciously. I would also like to thank my oldest friends James Corbett, Jess Chitty

and Christina Griffin who have cheered me on through the good times, and the

bad.

Lastly I would like to thank my fiancé Oliver Hill-Andrews who I met at the

beginning of these four years. His love and intellect are a continual inspiration to

me and have spurred me to complete this project.

5

Table of Contents

SUMMARY .............................................................................................................................. 3

ACKNOWLEDGMENTS ............................................................................................................... 4

TABLE OF CONTENTS ................................................................................................................ 5

1. CHAPTER I: INTRODUCTION ............................................................................................. 9

SUMMARY .............................................................................................................................. 9

1.1. THE NEED FOR RENEWABLE ENERGY .................................................................................... 9

1.1.1. Biofuels as a solution, for better or worse? ........................................................ 10

1.1.2. First generation biofuel ....................................................................................... 11

1.1.3. Lignocellulosic biomass and advanced second generation biofuel production.11

1.2. THE INDUSTRIAL BIOFUEL PRODUCER GEOBACILLUS THERMOGLUCOSIDANS ............................. 12

1.2.1. Geobacillus classification, environmental isolation and history ........................ 13

1.2.2. Genes and Genome ............................................................................................ 14

1.2.3. Industrial use of Geobacillus thermoglucosidans .............................................. 15

1.2.4. Advantages of Geobacillus thermoglucosidans as a biofuel producer .............. 16

1.3. CONTROL OF RESPIRATION IN BACTERIA .............................................................................. 18

1.3.1. Overview of the respiratory chain ....................................................................... 19

Glycolysis ................................................................................................................. 19

TCA cycle ................................................................................................................. 20

Oxidative phosphorylation ...................................................................................... 20

Oxidative phosphorylation in E. coli ............................................................... 21

ATP Synthase .......................................................................................................... 24

Oxygen independent respiration ............................................................................ 24

1.3.2. Regulators and sensors ...................................................................................... 26

ResDE ...................................................................................................................... 26

Fnr ........................................................................................................................... 26

ArcAB ....................................................................................................................... 27

1.4. REX IS A KEY REGULATOR OF RESPIRATORY AND FERMENTATION GENES IN GRAM-POSITIVE BACTERIA

................................................................................................................................................28

1.4.1. Discovery of Rex in Streptomyces ...................................................................... 28

1.4.2. Rex is a redox regulator in bacteria .................................................................... 29

1.4.3. The structure and function of Rex ...................................................................... 30

1.5. PROJECT AIMS ............................................................................................................... 35

2. CHAPTER II: MATERIALS AND METHODS ....................................................................... 36

2.1. CHEMICALS, REAGENTS, ENZYMES .................................................................................... 36

2.1.1. Chemicals and reagents ..................................................................................... 36

2.1.2. Enzymes .............................................................................................................. 37

2.1.3. Antibodies ........................................................................................................... 38

2.1.4. Solutions and buffers.......................................................................................... 38

2.2. STRAINS, VECTORS AND OLIGOS ........................................................................................ 40

2.2.1. Plasmids and expression vectors ....................................................................... 40

2.2.2. Bacterial strains .................................................................................................. 41

E. coli strains .......................................................................................................... 41

G. thermoglucosidans strains................................................................................ 41

2.2.1. Oligonucleotides ................................................................................................. 41

2.3. GROWTH MEDIA, STORAGE AND SELECTION ......................................................................... 44

6

2.3.1. Media .................................................................................................................. 44

2.3.2. Antibiotics and additives..................................................................................... 45

2.3.3. Growth and storage of bacterial strains ............................................................. 46

Growth of E. coli cultures and storage .................................................................. 46

Preparation of chemically competent E. coli by CaCl2 method............................ 46

Preparation of electrocompetent E. coli................................................................ 46

Growth of G. thermoglucosidans cultures and storage........................................ 47

Preparation of electrocompetent G. thermoglucosidans cells ............................ 47

2.4. DNA AND RNA MANIPULATION AND MOLECULAR CLONING .................................................... 48

2.4.1. DNA gel electrophoresis ..................................................................................... 48

2.4.2. RNA gel electrophoresis ..................................................................................... 48

2.4.3. DNA gel purification ............................................................................................ 48

2.4.4. General polymerase chain reaction (PCR).......................................................... 48

2.4.5. Inverse PCR ........................................................................................................ 49

2.4.6. Colony PCR ......................................................................................................... 49

2.4.7. DNA ligation ........................................................................................................ 50

2.4.8. Oligonucleotide annealing for construction of dsDNA fragments ...................... 50

2.4.9. Restriction digest ................................................................................................ 50

2.4.10. Dephosphorylation of DNA ............................................................................... 51

2.4.11. Phosphorylation of DNA primers ...................................................................... 51

2.4.12. Determination of DNA/RNA concentration ...................................................... 51

2.4.13. Gibson assembly .............................................................................................. 51

2.4.14. DNA sequencing ............................................................................................... 52

2.4.15. Generating an in-frame disruption strain using λ RED ..................................... 52

2.4.16. DNA purification ............................................................................................... 52

Small scale miniprep plasmid purification ......................................................... 52

Large scale midiprep plasmid purification ......................................................... 53

Chromosomal DNA extraction ............................................................................. 53

2.4.17. RNA Isolation and purification .......................................................................... 54

Cryogenic grinding method .................................................................................. 54

Rapid high throughput sonication method ......................................................... 54

2.4.18. DNA introduction into bacteria ......................................................................... 55

E. coli transformation........................................................................................... 55

Heat shock .................................................................................................... 55

Electroporation ............................................................................................. 55

G. thermoglucosidans transformation ................................................................ 56

Electroporation ............................................................................................. 56

Conjugation ................................................................................................... 56

2.4.19. Analysis of nucleic acids ................................................................................... 57

qRT PCR ................................................................................................................ 57

2.5. PROTEIN EXPRESSION AND PURIFICATION ........................................................................... 57

2.5.1. Protein expression .............................................................................................. 57

2.5.2. Protein purification ............................................................................................. 57

2.5.3. Nickel sepharose hand-made column ................................................................ 58

2.5.4. Concentrating Protein ......................................................................................... 58

2.5.5. Determining a protein concentration ................................................................. 58

Bradford assay........................................................................................................ 58

Qubit ........................................................................................................................ 58

2.6. SDS POLYACRYLAMIDE GEL ............................................................................................. 59

2.6.1. Sample preparation ............................................................................................ 59

2.7. WESTERN BLOT .............................................................................................................. 59

2.7.1. Semi-dry transfer ................................................................................................ 59

7

2.7.2. Washes and antibodies ...................................................................................... 60

2.7.3. ECL detection and development ......................................................................... 60

2.8. CHROMATIN IMMUNOPRECIPITATION (CHIP –SEQ) ................................................................ 60

2.9. ELECTROMOBILITY SHIFT ASSAY (EMSA) ............................................................................ 62

2.9.1. Generating radiolabelled DNA probe .................................................................. 62

2.9.2. Running EMSA gel .............................................................................................. 62

3. CHAPTER III: DEVELOPMENT OF TOOLS ........................................................................ 63

OVERVIEW ............................................................................................................................ 63

3.1. PRELIMINARY EVIDENCE FOR E. COLI TO G. THERMOGLUCOSIDANS INTERGENERIC CONJUGATION.64

3.1.1. E. coli strains ...................................................................................................... 64

3.1.2. Initial biparental mating protocol ....................................................................... 65

3.2. GROWTH MEDIA ............................................................................................................. 66

3.3. INVESTIGATING THE IMPORTANCE OF DNA METHYLATION ON CONJUGATION FREQUENCY ............. 67

3.3.1. Construction of S17-1 ∆dcm::apr donor strain using a REDIRECT approach .... 67

3.4. CONSTRUCTION OF A NON-REPLICATING CONJUGATIVE PLASMID .............................................. 71

3.5. DEVELOPMENT OF AND OPTIMISING A RAPID CONJUGATION PROTOCOL ..................................... 73

3.5.1. Conjugation temperature and time optimisation ............................................... 73

3.5.2. Recovery Period .................................................................................................. 74

3.5.3. Description and Comparison of Protocols .......................................................... 76

3.6. ADAPTING SEVA MODULAR SHUTTLE VECTORS .................................................................... 77

3.6.1. Testing conjugation frequency of pG1AK-sfGFP_oriT ......................................... 81

3.7. ATTEMPTS TO DEVELOP AN INTEGRATIVE CONJUGATIVE SYSTEM .............................................. 81

3.7.1. Integrative Vector ................................................................................................ 83

3.7.2. Genomic attachment site ................................................................................... 86

The orotate phosphoribosyltransferase (pyrE) gene as an attachment site ....... 86

pyrE gene disruption ....................................................................................... 87

The uracil phosphoribosyl-transferase gene (upp) as an attachment site.......... 90

upp gene disruption ........................................................................................ 90

3.8. DISCUSSION AND FUTURE WORK ....................................................................................... 92

4. CHAPTER IV: ANALYSING THE REX REGULON ............................................................... 95

OVERVIEW ............................................................................................................................ 95

4.1. THE GENETIC CONTEXT OF REX .......................................................................................... 96

4.2. IDENTIFICATION OF REX TARGETS BY CHIP-SEQ ANALYSIS ...................................................... 97

4.2.1. Construction of a 3xFLAG tagged rex strain ....................................................... 97

4.2.2. Optimisation of chromatin immunoprecipitation.............................................. 100

4.2.3. Chromatin immunoprecipitation of G. thermoglucosidans grown in liquid

cultures ....................................................................................................................... 101

4.2.4. ChIP-qPCR of predicted Rex target sites........................................................... 101

4.2.5. Chip-Seq and identifying the Rex regulon ........................................................ 102

4.2.6. Putative Rex targets involved in fermentation and anaerobic respiration ....... 110

4.2.7. New Rex targets ................................................................................................ 111

4.2.8. Rex binding site analysis .................................................................................. 114

4.3. CONSTRUCTION AND ANALYSIS OF A REX MUTANT IN G. THERMOGLUCOSIDANS ....................... 116

4.3.1. Initial attempts to isolate a Δrex mutant .......................................................... 116

4.3.2. Recombination of Δrex allele into G. thermoglucosidans ................................ 117

4.4. GENE EXPRESSION IN A ΔREX MUTANT ............................................................................. 118

4.4.1. Optimisation of growth conditions and RNA isolation ...................................... 118

4.4.2. Transcript analysis of WT and Δrex mutant strains .......................................... 119

4.4.3. General growth characteristics of the ∆rex mutant in liquid medium .............. 122

4.5. FERMENTATION ANALYSIS OF ΔREX MUTANT ..................................................................... 122

8

4.5.1. Growth and analysis of Δrex mutant in fermenters .......................................... 123

4.5.2. HPLC analysis ................................................................................................... 125

Metabolite analysis of the Δrex strain ................................................................ 125

4.6. FUTURE WORK AND DISCUSSION ..................................................................................... 129

4.6.1. Other regulators are likely to influence Rex target gene expression................ 130

4.6.2. Fermentation and the metabolome .................................................................. 131

5. CHAPTER V: DEVELOPMENT OF ORTHOGONAL REDOX -RESPONSIVE EXPRESSION

VECTORS ............................................................................................................................. 132

OVERVIEW .......................................................................................................................... 132

5.1. IDENTIFICATION OF POTENTIALLY ORTHOGONAL AND THERMOSTABLE REX REGULATORS ............ 132

5.2. CLONING AND OVEREXPRESSION OF CLOSTRIDIUM THERMOCELLUM REX PARALOGUES ............ 135

5.2.1. Design and overexpression of rex .................................................................... 135

5.3. DESIGN OF EMSA PROBES ............................................................................................ 137

5.4. DNA BINDING ANALYSIS OF REX PROTEINS USING EMSA.................................................... 139

5.4.1. C-Rex is responsive to NADH and NAD+ ........................................................... 141

5.5. DESIGN AND CONSTRUCTION OF REX-BASED GENETIC CIRCUIT EXPRESSION VECTORS ............... 143

5.5.1. Design of rex gene modular components and auto-regulating promoter ........ 144

5.5.2. Analysis of the auto regulating promoter controlling rex gene expression ...... 147

5.5.3. Design and analysis of promoter controlling gfp gene expression ................... 149

Design of Rex responsive promoters / operators .............................................. 149

Study of double and single ROP sites by expression analysis of gfp reporter. 152

5.5.4. Response of Rex expression vectors to oxygen limitation ............................... 153

5.5.5. Analysis of chromosomal Rex targets in response to oxygen limitation .......... 156

5.5.6. Complementation of the rex mutant ................................................................ 159

5.6. DISCUSSION AND CONCLUSIONS ..................................................................................... 161

6. CHAPTER VI: DISCUSSION ........................................................................................... 165

OVERVIEW .......................................................................................................................... 165

6.1. GENERAL DISCUSSION AND FUTURE DIRECTIONS ................................................................ 166

7. BIBLIOGRAPHY ............................................................................................................ 172

8. APPENDIX .................................................................................................................... 189

9

1. Chapter I: Introduction

Summary

Biofuels and other bio-based products are renewable, viable alternatives to

petroleum-based fuels and chemicals. The bacterium Geobacillus thermoglucosidans

has been utilised as a converter of lignocellulosic biomass to bioethanol. However,

to diversify products and optimise yields, innovation in genetic engineering tools is

badly needed. The metabolism of G. thermoglucosidans must be optimised and

redirected towards production of the required chemical. For this to be realised, not

only are improved genetic tools required, but an in-depth knowledge of the

regulators of gene expression is needed.

The sections that follow make the case for biofuels as an alternative to petroleum

and highlight the use of synthetic biology in aiding this transition process. The

Geobacillus genus, particularly Geobacillus thermoglucosidans, is then introduced as a

biofuel producer, including the advantages of the organism that have led to its

industrial application. An overview of how bacteria regulate respiration and

fermentation in response to environmental conditions follows, including the

mechanism by which the main regulators operate. A description of Rex, a redox

sensing regulator that is the focus of this thesis, is then given, followed by an

account of the aims of this project.

1.1. The need for renewable energy

Energy is humanity’s most important resource, its consumption continually

growing, yet the method of its production poses the greatest challenge of our time.

At present fossil fuels remain the main resource used to meet energy demands. The

impact of fossil fuel utilisation, however, is clearly environmentally damaging and

unsustainable (Armaroli & Balzani, 2011; WEF, 2016). A 2oC global temperature

increase limit has been set, a commitment intended to avoid the most serious effect

of climate change – such as the melting of the Greenland ice sheet and sea level

rises of 7 meters (Centre & Office, 2013). The international community has

recognised the need for change and governments and industry have made

10

investments in an array of alternative technologies, an investment increase of 80%

since 2000 (IEA, 2015). These are being developed and implemented, resulting in

diversification of the energy market, including wind power, solar energy,

geothermal heat, hydroelectricity and bioenergy. A global low-carbon energy target

has been set of 21% of the total energy supply by 2025 (IEA, 2015). Each energy

source has advantages and disadvantages associated with it, and each will be most

efficient when applied to particular applications. The niche biofuels occupy is as an

alternative to crude oil. Crude oil is a resource which is rapidly declining and

increasing in price as it becomes more difficult to extract (despite this the market

supply has steadily increased to drive economic growth) (Miller & Sorrell, 2014).

Biofuels can not only be renewable and potentially carbon neutral, but they also

represent a fresh commerce opportunity to replace the production and sale of

petroleum-based fuel and chemicals, often monopolised and linked to the

geographical location and natural resources of a country.

1.1.1. Biofuels as a solution, for better or worse?

Crude oil is the predominant resource used for production of both liquid

transportation fuels and a vast array of chemical products. The development of

biofuels, using microorganisms to convert renewable feedstocks, offers an

alternative, potentially superior in terms of both environmental impact and

renewability (Hill, Nelson, Tilman, Polasky, & Tiffany, 2006). When combusted,

biofuels also emit polluting gases, however they can be considered carbon neutral as

the feedstock is based upon renewable plant material. Plants absorb carbon dioxide

whilst growing and so the quantity released back into the atmosphere is equal to

that absorbed by the process of plant growth (Bogner et al., 2008). Although biofuel

feedstock is renewable and carbon neutral, several issues concerning the source and

exact type of feedstock exist, as this too can generate environmental and social

problems. Lignocellulosic biomass is often discarded as a waste material by the

agricultural and forestry industries, and makes a convincing candidate for use as a

second generation biofuel feedstock. Lignocellulosic biomass, however, is not

widely utilised by the current model microorganisms such as E. coli.

11

1.1.2. First generation biofuel

First generation biofuels, although utilising a renewable feedstock, have detrimental

social and environmental impact factors associated with production. Feedstocks

such as sugar and starches have been utilised for biofuel production due to the wide

range of microorganisms able to process them and the minimal pre-treatment

requirements. Negative implications of first generation biofuel production have

become evident (Sims, Taylor, Jack, & Mabee, 2008). Feedstocks such as sugar

cane or corn have become diverted from consumption as nutrition, for either cattle

or humans, towards biofuel production. This has repercussions for availability and

pricing of food, especially in the developing world. The resources needed to grow

the feedstock — such as fertilisers, land and water — are at risk of over

consumption, with potentially harmful environmental implications (Lopes, 2015).

The move towards alternative, second generation biofuels aims to alleviate many of

these concerns and their utilisation has recently become a focus of research

(Simmons, Loque, & Blanch, 2008).

1.1.3. Lignocellulosic biomass and advanced second generation

biofuel production

Second generation biofuels differ from first generation by the feedstock utilised in

their production. The alternative feedstock attempts to minimise any negative

aspects of first generation biofuels, but they tend to be organic waste materials that

are harder to break down. Substantial pre-treatment is usually needed which

includes multiple stages, requiring specialised equipment and energy use, impacting

overall energy yield and efficiency (Kumar, Barrett, Delwiche, & Stroeve, 2009;

Williams, Inman, Aden, & Heath, 2009). Lignocellulosic biomass, an abundant

potential feedstock, has been identified as being an alternative to first generation

feedstocks.

Lignocellulose biomass is composed of cellulose, hemicellulose (such as non-

cellulosic polysaccharides including xylans) and lignin (Abdel-Rahman, Tashiro, &

Sonomoto, 2010). The biomass can be found in several forms, such as crop

residues, municipal waste, wood and paper waste (John, Nampoothiri, & Pandey,

2007). The largest proportion of lignocellulose biomass is composed of cellulose,

made up of glycosidic bond D-glucose. Lignin surrounds cellulose fibres and is

12

composed of a complex matrix of aromatic alcohol bond linkages which make it

extremely hard to degrade (Abdel-Rahman et al., 2010). However, hemicellulose —

which is also sheathed around cellulose — is more easily hydrolysed. Hemicellulose

is composed of pentose (xylose) and hexose (glucose) sugars which can be used as

fermentation feedstock (Nieves, Panyon, & Wang, 2015). The proportion of

cellulose, hemicellulose and lignin differs between biomass source and pre-

treatment must be optimised accordingly.

The pre-treatment of lignocellulosic biomass involves initial mechanical

breakdown. This is followed by acidic chemical treatment used to degrade

hemicellulose polymers (Alvira, Tomás-Pejó, Ballesteros, & Negro, 2010). A costly

enzymatic step then follows to act upon cellulose and lignin, breaking it down and

making it accessible to microorganisms. Next, fermentation and the conversion of

the sugar monomers and dimers to fermentation products occurs (Saha, 2003).

Extraction of the biofuel and downstream processing follows. Each stage of pre-

treatment is optimised to maximise yield, but the process is demanding of both time

and energy, reducing the overall yield and therefore margin of profit. Industrially

relevant microorganisms must be capable of the co-utilisation of both pentose

(xylose) and hexose (glucose) sugars to be efficient lignocellulosic biomass to

biofuel converters. Although a variety of organisms exist that are able to utilise

lignocellulosic biomass such as fungi, they are not commercially viable for a

number of reasons. Not only must the organism possess the ability to utilise the

feedstock; it must also have the ability to produce and tolerate high concentrations

of a biofuel such as ethanol or butanol. Furthermore, the organism must produce

minimal amounts of unwanted products. A fast metabolism for quick product

turnaround and robustness for changes in temperature and pressure are also

essential attributes for a commercially viable organism.

1.2. The industrial biofuel producer Geobacillus thermoglucosidans

The Geobacillus genus has emerged as an important industrial strain, commercially

adopted for biofuel production (Taylor et al., 2009). It is a gram positive, rod

shaped bacterium characterised as facultative anaerobic and thermophilic, able to

ferment a variety of feedstocks, including sugars derived from lignocellulose and

13

municipal waste materials to products such as ethanol. The range of genetic

manipulation tools for use with this organism is underdeveloped, however. The

application of traditional genome editing techniques alongside synthetic biology

methods should quicken strain development, increasing yields and enabling

diversification of product production from ethanol to butanol, and indeed to other

high value commodities.

1.2.1. Geobacillus classification, environmental isolation and history

The Geobacillus genus is thermophilic and isolated from heated environments such

as compost heaps, hot springs, oil and gas wells. Geobacillus thermoglucosidans,

originally known as Bacillus thermoglucosidasius (Coorevits et al., 2012; Nazina,

Tourova, & Poltaraus, 2001), is the organism of focus throughout this thesis.

Therme, meaning heat, refers to the thermophilic nature of the bacterium, the

optimal growth temperature being between 55-65oC. The glucosidans component of

its name indicates its starch-hydrolyzing glucosidase activity, a key advantage of the

species as a biofuel producer. G. thermoglucosidans was first isolated from Japanese

soil and has been found to tolerate growth temperatures of between 42–69oC

(Suzuki et al., 1983). It was phenotypically analysed and originally categorised as

Bacillus thermoglucosidasius but, following the discovery of large numbers of other

thermophilic strains, and subsequent 16S rRNA gene sequence analysis, it was

transferred to a new genus, that of Geobacillus (Nazina, Tourova, Poltaraus, et al.,

2001). Further phenotypic analysis of multiple Geobacillus thermoglucosidasius strains

revealed that four were in fact of the species G. toebii. The remaining eight G.

thermoglucosidasius strain descriptions were amended and the name revised to G.

thermoglucosidans (Coorevits et al., 2012).

Prokaryotic organisms are divided into two domains, bacteria and archaea. The

bacterial domain is further subdivided into phyla including the Firmicutes which

are characterised as having a low C+G content (Bergey, 2009). Bacillales is an

order within this phylum, and the genus Geobacillus is a further subcategory to

which the species G. thermoglucosidans belongs, based on 16S rRNA gene sequence

analysis. The genus Geobacillus is part of a vast assortment of predominantly soil

dwelling organisms that contribute to the degradation of biomass and thereby play

an important role in the carbon cycle, maintaining a soil’s ability to support plant

14

life (Gougoulias, Clark, & Shaw, 2014). Each of these organisms, including G.

thermoglucosidans, occupies a specialist niche. Many soil characteristics such as

nutrient levels, temperature and oxygen concentration change dramatically

throughout a soil profile, giving rise to unique micro-ecosystems. The optimal

environment for G. thermoglucosidans may only arise infrequently; often the

bacterium will have to cope with more stressful environments.

During times of intolerable environmental stress, organisms can lay dormant until

circumstances improve. The genus Geobacillus achieves this by forming tough

endospores capable of resisting harsh conditions (Zeigler, 2014). When the

environment becomes more favourable, endospore reactivation occurs in stages.

Firstly, activation arises (triggered commonly by heating), germination shortly

follows with an increase in metabolic activity, finally leading to outgrowth and a

fully functional replicating cell.

G. thermoglucosidans are facultative anaerobic bacteria, able to respire in the absence

of oxygen. This is an important adaptation, as after the predominantly obligate

aerobic fungi have broken down complex organic matter at the surface of soil,

facultative anaerobes are integral for the continued decomposition within the depth

of soil, where low oxygen concentrations arise (Juo, 2003). For short periods

temperatures can reach upwards of 55oC due to the exothermic reactions and

decomposition of plant matter (such as in compost heaps), a temperature range G.

thermoglucosidans thrives at (Zeigler, 2014).

1.2.2. Genes and Genome

Many species of the Geobacillus genus have been sequenced and annotated, and are

stored on the NCBI genome database. These genomes include six G.

thermoglucosidans species strains (see Table 1.1). The median GC content of the G.

thermoglucosidans genome is 43.8%, and of a median length of 3.87 Mb in total. The

chromosome of G. thermoglucosidans is circular: in one strain the genome has been

found to contain two additional plasmids (Brumm, Land, & Mead, 2015).

The genome of C56-YS93 strain is unique in that it includes two circular plasmids

approximately 81 Kb and 20 Kb in size. This strain was also found to have a xylan

degradation cluster, complete with enzymes and transporters, not found in other

15

strains (Brumm et al., 2015). The C56-YS93 strain also contains nitrate respiration

clusters for reduction of nitrate to nitrite.

Strain Size (Mb) GC% Genes Proteins Isolation source Date

C56-YS93 3.99379 43.98 3992 3717 Hot Spring, USA 2011/06/10

DSM 2542 390714 43.70 3895 3667 Soil, Japan 2015/03/30

TNO-09.020 3.74024 43.90 3750 3521 Casein fouling sample from a dairy

factory, geographical location

unknown

2012/04/05

NBRC

107763

3.87116 43.70 3849 3644 Japan, specifics unknown 2014/04/12

W-2 3.89456 43.20 3919 3717 Oil reservoir, China 2016/06/03

GT23 3.69646 43.80 3666 3508 Casein pipeline, Netherlands 2016/05/25

Table 1.1 Genomes of G. thermoglucosidans species strains stored on NCBI database

(http://www.ncbi.nlm.nih.gov/genome/genomes/2405).

The G. thermoglucosidans strain 11955 was isolated and sequenced by NCIMB, a

microbiology and chemical company which collects industrially valuable

microorganisms. A wild type strain was selected by the second generation biofuel

producing company TMO Renewables Ltd/ReBio Technologies Ltd to undergo

strain development. It was chosen as a suitable strain for bioethanol production as it

can utilize C5 and C6 sugars.

1.2.3. Industrial use of Geobacillus thermoglucosidans

G. thermoglucosidans has been industrially adopted as a producer of bioethanol by

the British company TMO Renewables Ltd, and more recently ReBio Technologies

Ltd. It has the advantage over first generation bioethanol producers of being able to

convert cellobiose, xylitol and other lignocelluloses-based feed stocks into

http://www.ncbi.nlm.nih.gov/genome/proteins/2405?genome_assembly_id=172418

http://www.ncbi.nlm.nih.gov/genome/2405?genome_assembly_id=172419




http://www.ncbi.nlm.nih.gov/genome/2405?genome_assembly_id=276496


http://www.ncbi.nlm.nih.gov/genome/genomes/2405

16

fermentable sugars (Cripps et al., 2009). Naturally, the organism would convert

these into a range of products including acetate, lactate, formate and ethanol. Strain

development, however, has enabled flux to be diverted into ethanol production.

The ability of G. thermoglucosidans to utilise second generation feedstock, combined

with thermophilic associated properties, led to this strain being considered as a

strong candidate for industrial biofuel production. G. thermoglucosidans was named

as a most promising candidate for biofuel production, owing to its comparatively

high product tolerance and broad substrate utilisation range (Tang et al., 2009).

Although G. thermoglucosidans demonstrates many advantageous characteristics for

bioethanol production, the application of metabolic engineering has further

improved the strain. The main areas of research importance within industry have

been to increase ethanol yield and tolerance (Jingyu Chen, Zhang, Zhang, & Yu,

2015). Superior ethanol yield has been achieved by repressing alternative

fermentation pathways. The commercial strain TM242 has multiple-gene

knockouts, designed to divert metabolism towards ethanol production (Cripps et

al., 2009).

1.2.4. Advantages of Geobacillus thermoglucosidans as a biofuel

producer

The thermophilic characteristics of G. thermoglucosidans lend themselves to the

industrial environment. There are a number of reasons for this. Firstly, large scale

mesophilic fermentations require cooling, however thermophiles limit the need for

this energy intensive practice (Zeldes et al., 2015). Secondly, an increased rate of

metabolism and substrate conversion leads to a quicker product output and better

efficiency (Bhalla, Bansal, Kumar, Bischoff, & Sani, 2013). Finally, the risk from

contamination is minimised, as few microorganisms can tolerate such high

temperatures. In addition, as ethanol has a low boiling temperature, its removal is

facilitated by higher temperatures (Cripps et al., 2009; Lynd, van Zyl, McBride, &

Laser, 2005; Shaw et al., 2008).

TMO Renewables Ltd, a former company utilising G. thermoglucosidans as a biofuel

producer, used genetic manipulation tools to modify the metabolism of Geobacillus

thermoglucosidans to enhance ethanol production (Cripps et al., 2009). Naturally, G.

thermoglucosidans can produce ethanol, lactate, formate and acetate during anaerobic

17

respiration and fermentation. It does this to regenerate NAD+ from NADH. By

repressing alternative pathways not involved in the direct production of ethanol,

yields were enhanced. The commercial strain TM242 has three main mutations

designed to increase flux to ethanol production and decrease flux of alternative

fermentation products. TM242 has gene deletions in lactate dehydrogenase (LDH)

and pyruvate formate lyase (PFL) which led to a reduction in unwanted lactate and

formate production and an increase in ethanol yield (see Figure 1.1) (Cripps et al.,

2009). Gene knockout was achieved by use of homologous recombination. Gene

flanking regions were cloned into temperature sensitive vectors. Vectors were

transformed into G. thermoglucosidans via electroporation and integration was

induced by raising the temperature while applying kanamycin selection. The second

crossover event was encouraged by removing the kanamycin selection during

several rounds of sub culturing. The gene pyruvate dehydrogenase (PDH) was

Figure 1.1 Overview of major relevent fermentation metabolic pathways in G. thermoglucosidans. Industrial strain TM242 is an ldh and pfl mutant with an enhanced pdh

promoter to redirect the pathway to enhance ethanol yields. Abbreviations: LDH lactate dehydrogenase, PFL pyruvate formatelyase; PDH pyruvate dehydrogenase; ADH alcohol dehydrogenase; (Cripps et al., 2009).

18

upregulated by using a strong promoter (taken from the lactate dehydrogenase gene

which contains a Rex binding site). Upregulation of PDH was done to redirect

metabolic flux towards ethanol production. Recently the genome of the wild type

and the industrial strain have been sequenced, and gene annotation is underway.

Despite advancements in strain development for biofuel production, the techniques

used (although effective) are time consuming and limited. Genetic manipulation

tools are still hindering the rate of progress and a superior understanding of the

regulatory mechanisms controlling fermentation genes will help to enable the

precise control over metabolism and optimisation of industrial G. thermoglucosidans

strains (Kananaviciute & Citavicius, 2015).

1.3. Control of respiration in bacteria

Bacteria have evolved complex pathways to adapt to environmental fluctuations,

including to oxygen deprivation. To adapt, bacteria have developed environmental

sensors and genetic regulators for flexible respiration and energy generation (Green

& Paget, 2004). Oxygen is the most prevalent electron accepter during respiration,

however some bacteria can utilise others. Depending upon the environmental

conditions, bacteria will respire aerobically, anaerobically or in a fermentative way.

Some bacteria are obligate aerobes, unable to respire and grow in anoxic

environments. Others are facultative anaerobes, capable of utilising alternative

electron acceptors such as nitrate instead of oxygen. Obligate anaerobes also exist,

to which trace levels of oxygen in the environment are toxic. Bacteria such as these

are thought of as extremophiles now, but before biological oxygen production

during the Archean Eon they were the predominant life form on the planet

(Hamilton, Bryant, & Macalady, 2016). Many environmental niches exist where

facultative and obligate anaerobes thrive, breaking down organic compounds to

alcohols and organic acids to produce energy. Facultative anaerobes in particular

face environments where oxygen concentrations fluctuate regularly. These bacteria

must be capable of rapidly shifting from utilising oxygen as an electron acceptor to

another energy generating pathway.

19

1.3.1. Overview of the respiratory chain

Respiration is an essential process for the production of adenosine triphosphate

(ATP), the energy currency of the living cell. Aerobic respiration can be split into

three main stages: carbohydrate catabolism, the citric acid (TCA) cycle and electron

transport. During glycolysis a sugar molecule is broken down to form ATP and

pyruvate which then enters the TCA cycle. Acetyl-CoA is subject to a series of

oxidation reactions that reduce NAD+ to NADH, which is utilised as an electron

donor for electron transport chain. Electron transport involves a series of oxidation

and reduction reactions that transfer electrons between membrane-associated

donors and acceptors, the function of which is to create an electrochemical gradient

created by hydrogen ions to produce a transmembrane proton motive force (PMF)

that is used for ATP synthesis (Marreiros et al., 2016).

Respiration can occur under two main conditions. Aerobic respiration occurs in

oxygen-abundant environments, whilst anaerobic respiration occurs in oxygen-

limited or anoxic conditions. During aerobic respiration oxygen is the ultimate

electron acceptor, resulting in the production of water (Borisov & Verkhovsky,

2015). Facultative anaerobes are able to transition from aerobic to anaerobic

respiration by utilising a complex regulatory network that senses and responds to

changes in environment (Richardson, 2000). By utilising alternative electron

acceptors, they are able to generate energy most efficiency depending upon the

abundancy of electron acceptors in the environment.

Glycolysis

Glycolysis is an oxygen independent process that occurs in the cytoplasm of

bacteria and is the first stage of ATP production. It is the metabolic pathway

responsible for converting carbohydrates, such as glucose, to pyruvate through a

series of reactions, releasing energy. During the process a six carbon glucose

molecule, over the course of nine reactions, is broken up into two molecules of

three carbon pyruvate. Pyruvate is utilised in subsequent metabolic pathways to

generate additional energy, including during either fermentation or the TCA cycle.

Along with the pyruvate molecules and ATP, two molecules of NADH are also

produced (see Figure 1.2). NADH molecules act as electron carriers and are re-

oxidised by entering oxidative phosphorylation, where ATP synthesis occurs.

20

TCA cycle

The TCA cycle (commonly referred to as the Krebs cycle after the 1953 Nobel Prize

winning scientist Hans Krebs [1900–1981], or alternatively known as the citric acid

cycle) follows on from glycolysis and is made up of a series of eight reactions. Once

pyruvate is converted into acetyl-CoA, it feeds into the TCA cycle and undergoes

reactions that yield reduced electron carriers NADH and FADH (Spaans,

Weusthuis, van der Oost, & Kengen, 2015). ATP and GTP are also produced along

with CO2. NADH is the most abundant product of the TCA cycle, which is used to

help form the PMF, which is subsequently used to generate ATP in the oxidative

phosphorylation. The TCA cycle also produces many intermediates which are

important molecules for other metabolic pathways and the production of structural

components. Each of the eight reactions is catalysed by a specific enzyme that is

precisely regulated by the cell, ensuring energy is metabolised in the most efficient

way possible.

Oxidative phosphorylation

The majority of energy released during respiration in glycolysis and the Krebs cycle

is locked in the coenzyme/electron acceptor NADH. During oxidative

phosphorylation this energy is utilised to produce ATP, and is the most efficient

and abundant method of production. Many different mechanisms of generating a

PMF for ATP synthesis exist, but all pass electrons from a donor, down the

electron transport chain, releasing energy in manageable increments to produce a

Glucose + 2 ATP + 2 NAD+ + 2 pi -->

2 Pyruvate + 2 ATP + 2 NADH + 2 H+

Figure 1.2 Glycolysis reaction. One glucose molecule produces two ATP and two NADH molecules plus two hydrogen ions during glycolysis. The pyruvate molecules are used in the TCA cycle and NADH feeds electrons into oxidative phosphorylation to generate a PMF.

21

proton motive force coupled to ATP synthesis. Electrons are carried from proteins

with low reduction potential, to those with higher reduction potential, and

ultimately to a terminal electron acceptor, which is oxygen during aerobic

respiration (Borisov & Verkhovsky, 2009).

Oxidative phosphorylation is a universal metabolic pathway; within prokaryotes

this process is located in the inner membrane, whilst in eukaryotes it occurs in

mitochondria. An overview of the electron transport chain in E. coli will facilitate a

good understanding of the mechanisms by which oxidative phosphorylation leads

to energy production.

Oxidative phosphorylation is conducted by the electron transport chain, made up of

a series of membrane bound dehydrogenases that reduce the quinone pool, and

terminal reductases that can oxidise it. The chain components are arranged so that

energy is released by a series of oxidative/reductive reactions to create a proton

motive force (PMF) coupled to the production of ATP. ATP synthases (known as

complex V) use the energy generated by the PMF to form the bonds needed for

ATP synthesis. Complexes can be comprised of numerous types of prosthetic

groups (hemes, flavins, iron-sulfur clusters and copper clusters) depending upon the

most energetically efficient method of electron transport given the circumstances

(Gottfried Unden, Steinmetz, & Degreif-Dünnwald, 2014). Gene regulators control

the expression of proteins involved in the electron transport chain, in response to

environmental conditions and respiration type (aerobic or anaerobic).

Oxidative phosphorylation in E. coli

E. coli has a branched respiration chain, with multiple dehydrogenases and terminal

reductases having been described, which enable multiple electron donors and

acceptors to be utilised by the organism (see Table 1.2). Electron acceptors include

oxygen during aerobic respiration, which is the most energetically efficient, or

alternatively nitrate or fumarate can be used (G Unden & Bongaerts, 1997). Each

enzyme has an associated redox potential, which increases along the transport

chain as electrons are transferred.

22

Electron donors dehydrogenases quinone pool reductases Electron

acceptors

NADH NADH dehydrogenase I Menaquinone Cytochrome bo Oxygen

Succinate NADH dehydrogenase II Ubiquinone Cytochrome bd Nitrate

Glycerophosphate Cytochrome c Dimethylmenaquinone Fumarate

Formate Succinate dehydrogenase

Hydrogen Glycerophosphate

Pyruvate Dehydrogenase

Lactate

Table 1.2 Lists of the most prevalent enzymes and electron donors/acceptors in E.coli (G Unden &

Bongaerts, 1997).

A dehydrogenase is the first protein in the electron transport chain. Many

alternative dehydrogenases exist that are able to accept electrons from a variety of

donors such as lactate and succinate. However, usually an NADH dehydrogenase

accepts electrons from NADH and passes them to the quinone pool (see Figure

1.3). In E. coli there are two types of NADH dehydrogenases (I and II), which are

expressed under different environmental conditions (Yagi, 1993). NADH

dehydrogenase I (prosthetic groups FMN, FeS) is a large protein complex that

spans the membrane, and is able to pass hydrogen ions into the intermembrane

space, contributing to the proton motive force (PMF). It also catalyses the reduction

of ubiquinone to ubiquinol and is expressed during times of oxygen starvation.

NADH dehydrogenases II (prosthetic group FAD) is a single subunit enzyme and

does not span the membrane, and therefore is unable directly to contribute to the

PMF. It also catalyses the reaction of ubiquinone to ubiquinol, but it is expressed

when oxygen is abundant. During conditions when NADH dehydrogenase II is

most highly expressed, the terminal reductase cytochrome bo is also upregulated, a

membrane spanning protein capable of generating a PMF. This compensates for the

23

downregulation of NADH dehydrogenases I and allows the cell to recycle NADH

to NAD+ whilst maintaining a PMF during both aerobic and anaerobic conditions.

In addition to NADH, intermediates of the TCA cycle such as formate and

succinate are also able to act as electron donors. Succinate dehydrogenase catalyses

the reaction of succinate to fumarate and ubiquinone to ubiquinol, but does not

contribute to the PMF directly. Glycerophosphate dehydrogenase is also an

alternative dehydrogenase utilised during anaerobic respiration (X. Zhang,

Shanmugam, & Ingram, 2010).

Once electrons have been donated to a dehydrogenase they are passed to the

quinone pool. Ubiquinone, ubiquinol, menaquinol and menaquinone make up the

quinone pool in E. coli and link the transfer of electrons from the dehydrogenases to

terminal reductases. Each of the quinones are most abundant during particular

growth conditions and are able to carry electrons or hydrogen ions and are small

mobile molecules. Ubiquinone is most abundant during aerobic respiration,

whereas menaquinone and dimethylmenoquinone dominate during anaerobic

conditions (G Unden & Bongaerts, 1997). Due to the midpoint potentials of the

quinone, each is only able to function with particular electron donors and

acceptors. Ubiquinone is used when oxygen is the electron acceptor, whereas

menaquinol are optimum for use with a nitrate terminal electron acceptor

(Gottfried Unden et al., 2014). Menaquinone is most optimal for DMSO and

fumarate electron donors (G Unden & Bongaerts, 1997).

In E. coli there are two main terminal oxidases: cytochrome bo, present at high levels

during oxygen abundance in aerobic conditions, and cytochrome bd, expressed most

NADH --> NAD+ + 2H+ + 2e- + ½ O2 --> H2O

Figure 1.3 The overall reaction of the electron transport chain. One NADH molecule

contributes two electrons, whose combined energy can transport ten protons to create a PMF. The electrons are eventually transferred to oxygen and combine with two hydrogen ions to form

water.

24

during periods of oxygen starvation (Borisov, Gennis, Hemp, & Verkhovsky,

2011). Cytochrome bd oxidises menaquinol but is unable to pump protons to

contribute to the PMF (Borisov, Gennis, Hemp, & Verkhovsky, 2011). It does

however have a higher affinity for oxygen then cytochrome bo and can transfer

electrons to produce water during microaerobic respiration (Borisov & Verkhovsky,

2015; Jasaitis et al., 2000). Cytochrome bo, however, does contribute to the PMF as

it contains prosthetic groups Heme O3 and CuB (also Heme b, although this does

not help generate a PMF) and oxidises ubiquinone in aerobic conditions, utilising

oxygen to produce water (Borisov et al., 2011).

ATP Synthase

The final stage of respiration is ATP synthesis which uses the electrochemical

gradient, created by the electron transport chain, to drive ATP synthesis. The F0F1

ATP synthase enzyme spans the membrane and allows hydrogen ions to flow down

a concentration gradient, thereby capturing energy and using it to form the bonds

needed for ATP synthesis. The F1 subunit comprises the catalytic site where ATP

synthesis occurs (using ADP and phosphate) (Trchounian, 2004; Weber & Senior,

2003). The F0 subunit makes up the section which spans the membrane, containing

a channel running though it for hydrogen ions to pass through. Variations of ATP

synthases are also used to catalyse the reverse reaction to enable other processes

such as movement by flagellum (Trchounian, 2004; Weber & Senior, 2003).

Oxygen independent respiration

There are two forms of oxygen independent energy production that can be engaged

depending upon environmental conditions: anaerobic respiration and fermentation.

In contrast to aerobic respiration, where oxygen serves as the final electron

acceptor, anaerobic respiration uses an alternative electron acceptor such as nitrate.

In this case the electrochemical gradients are generated using a specialised

membrane-associated electron transport chain. Fermentation, however, occurs

without the use of external electron acceptors: rather, an internally generated

metabolite such as pyruvate is used as an electron acceptor and is reduced to

energy-rich products such as lactate, butanol and ethanol (Härtig & Jahn, 2012).

The main purpose of this fermentative process is to create redox balance by the

25

process of recycling NADH back into NAD+, allowing continued ATP production

by substrate-level phosphorylation. Many fermentation products are toxic at high

concentrations and are excreted by the cell.

Bacteria can be categorised by their ability to adapt to changes in oxygen

abundance and their method of energy production. Obligate aerobes require oxygen

for oxidative phosphorylation and cannot grow in its absence (although they can

often adapt their aerobic respiratory pathways to account for variation in oxygen

levels). Facultative anaerobic bacteria are able to respire in the presence of oxygen,

and adapt to the absence of oxygen by switching to anaerobic respiratory and/or

fermentative pathways. Obligate anaerobes can only grow anaerobically and often

perish in oxygen-rich environments.

It is the concentration of oxygen and the substrates available that determine the

state of respiration within the cell of a facultative anaerobic bacterium. Facultative

anaerobes are therefore able to quickly adapt to changing conditions by directly

sensing the amount of oxygen or indirectly, by sensing for example redox signals.

Thus, they can control gene expression in response to a variety of environmental

conditions, allowing them to inhabit a diverse range of territories and adapt to

changing conditions (Morris & Schmidt, 2013).

The Geobacillus genus is made up of a combination of both obligate and facultative

anaerobes. Soil inhabiting microorganisms such as these face constantly fluctuating

oxygen concentrations. During the decomposition of organic matter within the

depths of soil, oxygen can become scarce. For this reason, facultative anaerobic

bacteria are integral to the continued recycling of carbon. After obligate aerobes at

the surface of soil have secreted enzymes that break down the complex organic

matter of wood and leaves (Juo, 2003), bacteria such as G. thermoglucosidans are able

to continue the degradation of biomass beneath soil by respiring not only

aerobically, but also surviving oxygen limitation by switching to fermentation or

anaerobic respiration. The sensing of oxygen levels and the switch between

respiration types involves a complex regulatory network.

26

1.3.2. Regulators and sensors

Control of metabolic flux and the transition between aerobic, anaerobic respiration

and fermentation is regulated by a number of sensors and regulators that influence

expression of important genes. The sensors and regulators involved form

interconnected networks that allow for a high degree of control over respiration in

response to a variety of environmental factors. This ensures that the cell generates

energy in the most efficient way possible, in relation to growth conditions,

switching off pathways that are unnecessary. There are many sensors and regulators

associated with changes in respiration including ArcAB in E. coli, ResDE in B.

subillis, Fnr in both organisms, and Rex in Gram positive bacteria. It is important to

understand these regulators, especially when using an organism for practical

applications such as biofuel production, so that control of key genes can be

engineered for optimal yield.

ResDE

The two component sensor kinase and response regulator ResD-ResE system in B.

subtilis is a key regulator of anaerobic respiration (Durand et al., 2015; Geng, Zuber,

& Nakano, 2007). Fnr expression is controlled by the ResD-ResE system in B.

subtilis, in response to oxygen and nitric oxide levels. ResE has the ability to sense,

via a PAS domain, both oxygen and nitric oxide. ResE then modulates ResD

activity via phosphorylation, by either phosphorylating or dephosphorylating ResD

(Baruah et al., 2004). The ResD-ResE system controls the switch to alternative

terminal electron acceptors, from oxygen to nitrate. It does this by regulating genes

such as cytochrome bf complex, ctaA haem biosynthesis, and the regulator Fnr

(Durand et al., 2015; Esbelin, Armengaud, Zigha, & Duport, 2009).

Fnr

The Fnr (Fumarate and Nitrate reductase) regulator senses oxygen in the

environment directly (Mazoch & Kucera, 2002). The Fnr protein is expressed by

both E. coli and B. subtilis and is activated under anaerobic conditions. Fnr has

combined sensory and regulatory activities, unlike the ArcAB two component

system in which they are separate, and the crystal structure of Fnr from Aliivibrio

fisheri has been solved recently (Volbeda, Darnault, Renoux, Nicolet, & Fontecilla-

27

Camps, 2015). Fnr contains an oxygen-sensitive iron-sulphur cluster (4Fe-4S) that is

required to allow it to bind DNA target sequences and thereby control gene

expression (Lazazzera, Beinert, Khoroshilova, Kennedy, & Kiley, 1996; Mettert &

Kiley, 2005). In the absence of oxygen, the E. coli 4Fe4S cluster binds via cysteine

residues to promote the formation of a protein dimer which can bind to DNA

(Tolla & Savageau, 2010). However in the presence of oxygen the dimerization

bonds are broken, preventing DNA binding (Lazazzera et al., 1996). In B. subtilis,

however, the Fnr protein forms a constant dimer independent of oxygen and

evidence suggests it binds to ResD, indicating that the mechanism of activation is

different to that in E. coli (Härtig & Jahn, 2012). In B. subtilis, Fnr positively

regulates genes involved in the transition between aerobic and anaerobic respiration

and fermentation, including: narGHJI operon that encodes nitrate reductase; arfM,

an anaerobic modulator; and narK, a nitrite extrusion transport protein (Nakano,

Dailly, & Zuber, 1997; Reents et al., 2006). Fnr has also been found to auto-

regulate as well as regulating the expression of other regulators such as ArcA in E.

coli (Reents et al., 2006).

ArcAB

E. coli is a facultative bacterium that can use nitrate as an alternative electron

acceptor instead of oxygen. The ArcAB two component system in E. coli regulates

the transition from aerobic to anaerobic respiration in E. coli, allowing it to use

nitrate as an alternative electron acceptor, and the transition involves initially the

use of alternative oxygen-dependent terminal oxidases (Alexeeva, Hellingwerf, &

Teixeira de Mattos, 2003). The ArcAB system is comprised of the response

regulator ArcA, along with ArcB, a membrane associated sensor

kinase/phosphatase (X. Liu & De Wulf, 2004; Loui, Chang, & Lu, 2009). ArcB is a

multi-domain protein with a signal sensor PAS domain that is thought to sense the

redox state of the quinone pool in the cell membrane through the reversible

formation of an intermolecular disulphide bond that occurs between two subunits

(Bekker et al., 2010). During aerobic respiration the majority of the quinone pool is

in the oxidised form, and it has been proposed to accept electrons from two cysteine

residues in the PAS domain of ArcB, leading to the formation of a disulphide bond

that inhibits its activity (Malpica, Franco, Rodriguez, Kwon, & Georgellis, 2004).

28

During oxygen limitation (microaerobic growth), the accumulation of the quinone

pool in a reduced state prevents the formation of a disulphide bridge, giving rise to

large structural changes which activate the auto kinase activity of ArcB. This leads

to the phosphorylation of ArcA and the binding to DNA to repress genes involved

in aerobic respiration and activate those important for anaerobic respiration, such

as pflA (X. Liu & De Wulf, 2004).

1.4. Rex is a key regulator of respiratory and fermentation genes in

Gram-positive bacteria

Rex is a redox sensor that controls the expression of numerous genes involved in

energy generation, such as fermentation genes, in many Gram-positive organisms

including strict anaerobes. The general model for the action of Rex is that it acts as

a repressor during growth that is not limited by terminal electron acceptor

availability, conditions in which the cellular NADH to NAD+ ratio is low on

account of the rapid recycling of NADH. During times of oxygen limitation, which

causes NADH levels to increase, Rex binds directly to NADH with high affinity

and disassociates from DNA, thereby allowing the transcription of target genes

(Brekasis & Paget, 2003). Rex is therefore instrumental in allowing the cell to sense

and respond to redox stress caused by changes in the rate of respiration and

fermentation caused, for example, by oxygen limitation.

1.4.1. Discovery of Rex in Streptomyces

Originally Rex was discovered in Streptomyces coelicolor and identified as a repressor

of cytochrome bd terminal oxidase operon (cydABDC; Brekasis & Paget, 2003).

The cydABDC operon is controlled by two promoters: cydp2 is constitutive and

cydp1 is induced during hypoxia. Mutagenesis of the cydp1 region revealed an

operator site (sequence 5’-TGTGAACGCGTTCACA-3’), which was designated

ROP (Rex operator) (see Figure 1.4). This was used as a template to identify other

ROP sites within the genome. The nuoA-N and SCO3320-hemACD operons were

found to contain upstream ROP sites. Interestingly, the first gene in the SCO3320-

hemACD transcriptional unit encoded a potential DNA-binding protein which

when overexpressed produced a protein that bound to the cydp1 ROP site (Brekasis

& Paget, 2003) and acted as a repressor. Subsequent deletion of the SCO3320 gene

29

resulted in the constitutive expression of the cytochrome cydA gene, verifying its

control of expression at this promoter site.

Figure 1.4 Map of cydABC operon in S. coelicolor (Brekasis & Paget, 2003; Sickmier et al., 2005).

The amino acid sequence of SCO3320 was predicted to include a Rossmann fold,

which is associated with the binding of pyridine nucleotides such as NADH.

Strikingly, the ability of SCO3320 protein to bind to DNA was inhibited by

NADH, whereas the addition of NAD+ to the assay counteracted the NADH

effects, allowing SCO3320 protein to bind DNA and repress gene transcription

again. The conclusion was therefore that the SCO3320 protein is able to detect the

ratio of NADH:NAD+ in the cell and control gene expression accordingly; it was

consequently named Rex (for redox regulator) due to its ability to detect redox

poise.

1.4.2. Rex is a redox regulator in bacteria

Following its initial discovery in S. coelicolor, many Rex orthologues have been

recognised in Gram-positive bacteria, including facultative (Geobacillus) and obligate

(Clostridia) anaerobes (Clostridia; Wietzke & Bahl, 2012) and facultative anaerobes

(Härtig & Jahn, 2012). The Rex binding site (ROP; TTGTGAnnnnnnTCACAA) is

remarkably well conserved across a number of taxonomic groups (Zheng et al.,

30

2014) and Rex has been identified as a key redox sensor and regulator of

fermentation genes in Gram-positive organisms (Härtig & Jahn, 2012; Sickmier et

al., 2005b; Wietzke & Bahl, 2012). A rex mutant of an industrial biofuel producer,

Clostridium acetobutylicum, has been studied and it was demonstrated that it

produced lesser amounts of hydrogen and acetone, but higher amounts of ethanol

and butanol compared to wild type (Wietzke & Bahl, 2012). It was discovered that

the adhE gene was over expressed in the C. acetobutylicum Δ rex and therefore that

Rex was of key importance to the solventogenic shift. The Rex gene was found to

be located just upstream of an operon, including genes vital for conversion of

acetyl-CoA to butyryl-CoA (see Figure 1.5), and many Clostridium species also have

similar gene arrangements, where rex is found in close proximity to carbon

metabolism and fermentative genes (Wietzke & Bahl, 2012).

The further investigation of and understanding of Rex is key to optimisation and

engineering of industrial biofuel-producing bacterial strains. The study of Rex will

provide a better understanding of the optimum bioreactor conditions to induce the

genes needed for biofuel production, and may also lead to the discovery that rex

regulates several genes important for biofuel production, which were previously

unknown.

1.4.3. The structure and function of Rex

Although the structure of G. thermoglucosidans Rex has not been solved, many others

have been. X-ray crystal structures are available for Rex proteins from Thermus

rex crt bcd etfB etfA hbd

Figure 1.5 The Rex gene and the crt-bcd-etfA/B-hbd (bcs) operon.

31

aquaticus (T-Rex) the identical Thermus thermophiles (T-Rex), B. subtilis (B-Rex),

Streptococcus agalactiae (S-Rex), and Thermoanaerobacter ethanolicus (RSP) (Protein

Data Bank [PDB] accession number 1XCB, 2DT5, 2VT2, 2VT3, 3IKT, 3IKV,

3IL2, 3KEO, 3KEQ and 3KET) (McLaughlin et al., 2010a; Nakamura, Sosa,

Komori, Kita, & Miki, 2007; Sickmier et al., 2005a; Wang, Ikonen, Knaapila,

Svergun, Logan, & von Wachenfeldt, 2011; Zheng et al., 2014). Of particular use is

the thermophilic Thermus aquaticus Rex x-ray structures, including it bound to

NADH (2.9 A resolution), NAD+, and DNA together with NAD+ (Sickmier et al.,

2005b; Wang, Ikonen, Knaapila, Svergun, Logan, & Wachenfeldt, 2011). These

structures have helped to establish which protein residues are important for DNA

and NADH interactions and the mechanism of redox sensing. G-Rex has a 40%

identity to T-Rex (see Figure 1.6) and binds to the same consensus DNA rex

operator (ROP) site. T-Rex, like all known Rex proteins, consists of three main

domains and naturally forms a homodimer. The C-terminus of Rex comprises of a

Rossmann fold that can bind NADH (Sickmier et al., 2005b; Wang, Bauer,

Rogstam, Linse, Logan, & von Wachenfeldt, 2008). In the absence of NADH, Rex

binds to DNA via an N- terminal winged helix-turn-helix motif, consisting of four

α-helices and two β-strands (Sickmier et al., 2005). The NADH binding protein

residues of T-Rex are located within the centre of the entire protein sequence. As

Figure 1.7 illustrates, the C-terminal domain of Rex is comprised of four α-helices

and seven β-strands, and forms a Rossmann fold. The final α-helix is domain-

swapped, interacting with the α-helix on the opposing subunit, and inserting

between the two domains to aid dimer formation (Sickmier et al., 2005). Rex has

been found to not only bind NADH, but also NAD+, and is sensitive to the ratio of

NADH/NAD+ within the cell (Brekasis & Paget, 2003; Sickmier et al., 2005).

Although Rex can bind NADH, it does not re-oxidise it: Rex merely senses redox

poise. Any effect on cellular redox state is indirect and via the regulation of gene

expression.

The binding of Rex to NADH causes a change in its conformation, inducing

dissociation and preventing interaction with DNA ROP sites. NADH/NAD+

molecules bind to Rex at the dimer interface and induce the rotation of the Rex

homodimer (Figure 1.8) (Mclaughlin et al., 2010). The NADH-bound Rex

conformation is “closed” and the positions of key residues central in DNA

32

interactions result in the inability of Rex to fit into the major groove of DNA, and

critically to form hydrogen bonds (McLaughlin et al., 2010). However, when

NADH dissociates from the T-Rex triggers 43o rigid body rotation of entire subunit

and 22Å movement in WH domain allowing DNA binding at ROP sites.

T-Rex 1 -----MKVPSAAISRLITYLRILEELEAKGIHRTSSEQLAEEAQVTAFQV

G-Rex 1 MNNEQPKIPQATAKRLPLYYRFLKNLHASGKQRVSSAELSEAVKVDSATI

consensus 1 ..... *.* * ** * * * * * * * ** * * * .

T-Rex 46 RKDLSYFGSYGTRGVGYTVPVLKRELRHILGLNRRWGLAIVGMGRLGSAL

G-Rex 51 RRDFSYFGALGKKGYGYNVNYLLSFFRKTLDQDEITEVALFGVGNLGTAF

consensus 51 *.* **** * .* ** * * *. * .*. *.* **.*

T-Rex 96 ADYPGFG-ETFELRGFFDVDPEKIGKPVGRGVIEPMEALPQRVPGRIEIA

G-Rex 101 LNYNFSKNNNTKIVIAFDVDEEKIGKEVGGVPVYHLDEMETRLHE-GIPV

consensus 101 * . . **** ***** ** . .. . *. .

T-Rex 145 LLTVPREAAQEAADQLVRAGIKGILNFAPVVLEVPKEVAVENVDFLAGLT

G-Rex 150 AILTVPAHVAQWITDRLVQKGIKGILNFTPARLNVPKHIRVHHIDLAVEL

consensus 151 . . . **

T-Rex 195 RLAFFILN-

G-Rex 200 QSLVYFLKN

consensus 201 . * .

Figure 1.6 Protein sequence alignment of Thermus aquaticus (T-Rex) Geobacillus

thermoglucosidans (G-Rex). Sequences share a 40% identity. The N-terminus of

Rex comprises of a winged helix-turn-helix motif. The C-terminus includes a

Rossman fold. The NADH binding triad includes R16, Y98, D188. R16 and

D188 also form the salt bridge that stabilizes Rex-ROP complex. Residue 47 is integral for ROP binding specificity.

16

47

98

188

33

The ROP site is a highly conserved 18 bp palindromic sequence present at promoter

regions controlled by Rex. Rex interacts with both the major and minor groove

through its winged helix-turn-helix domains, and the spacing and orientation of

these is therefore of major importance. Interestingly, while ROP sites are generally

highly conserved across bacteria, Clostridia appear to deviate, having a binding

consensus of: TTGTTAANNNNTTAACAA, compared to Actiobacteria and

Bacilli; TTGTGAANNNNTTCACAA (D. a. Ravcheev et al., 2012). This

difference in binding motif corresponds to substitutions in protein residues found in

the DNA recognition α -helices. Clostridia has been the only Rex to have a

substituted Gln47, an important hydrogen-bond-forming residue that interacts with

guanine7 ROP base. Actionobacteria have a Lys47, and Bacilli a Arg47, both of

which are positively charged amino acids, essential for hydrogen bond formation

(Zhang et al., 2014). The Rex regulon varies across taxonomic groups from

Figure 1.7 T-Rex structure of dimer. The Rossmann fold is a NADH binding domain and the wiggled helix is a DNA binding domain (Mclaughlin et al., 2010; Sickmier et al.,

2005).

34

including genes involved in energy metabolism, NAD(P) biosynthesis,

fermentation, nitrogen and sulphur reduction and central carbon metabolism (D. a.

Ravcheev et al., 2012). Harnessing and controlling Rex and the role it plays in

controlling such genes would be of use in the industrial production of biofuels and

other biocommodities.

Figure 1.8 Rex binding to adh ROP site. During periods of oxygen limitation NAD+ builds up

and interacts with Rex, making it disassociate from the ROP site and allowing expression of adh.

35

1.5. Project aims

The aims of this study center on understanding the role of Rex, a regulator of

fermentation. To investigate Rex, it was aimed to construct two strains using two

alleles: a strain in which Rex was tagged with a FLAG epitope, and a strain where

rex was deleted. The rex-FLAG allele would be integrated into the native site,

replacing the original rex gene. The tag would allow the investigation of the G.

thermoglucosidans Rex regulon using chromatin immunoprecipitation techniques,

combined with high-throughput DNA sequencing. The rex deletion mutant allele

would also be integrated into the native locus and the resulting strain analysed for

changes in the transcriptome using qRT-PCR and changes in the metabolome (e.g.

fermentation products such as ethanol, lactate and formate) using HPLC.

This study also aimed to develop new genetic manipulation tools for the improved

transformation of G. thermoglucosidans. This was to include the development of

vectors and E. coli strains for intergeneric conjugation as well as the optimisation of

protocols.

Finally, it was aimed to apply the basic understanding of the Rex system towards

the construction of synthetic genetic regulatory systems to manipulate the

metabolism of G. thermoglucosidans. The work will be proof of concept,

demonstrating the ability to control gene expression by using a potentially

orthoganol Rex.

Ultimately the study aims to contribute to the understanding of fermentation

regulation in these industrial organisms and to develop genetic engineering tools

that can be implemented to demonstrate the potential for metabolism engineering

for the efficient production of biofuels and other biocommodities.

36

2. Chapter II: Materials and methods

2.1. Chemicals, reagents, enzymes

2.1.1. Chemicals and reagents

Acrylamide solutions (Severn Biotech Ltd)

Agarose powder (Melford Laboratory)

Ammonium persulphate (Sigma)

Ampicillin (Melford)

Apramycin (Duchefa Biochemie)

Bradford Reagent (Sigma)

Bromophenol blue (Amersham Biosciences)

Casamino acids (Difco)

Chloramphenicol (Melford)

Chloroform (Fisher)

Dimethylsulphoxide (DMSO) (BDH)

Dithiothreitol (DTT) (Melford)

dNTPs (New England Biolabs)

Glycogen (Fisher)

Hepes (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) (Fisher)

Isopropyl-β-D-thiogalactopyranoside (IPTG) (Melford)

Isoamyl alcohol (Sigma)

Kanamycin (Melford)

N, N-dimethyl-formamide (Fisher)

N-Tris(hydroxymethyl)methyl-2-aminoethane sulfonic acid (TES) (Fisher)

37

Nutrient agar (Difco)

Phenol (Fisher Scientific)

PEG 1000 (Polyethylene glycol) (BDH)

Phenol (Fisher Scientific)

Phenylmethylsulfonyl Fluoride (PMSF) (Sigma)

Radionuclides ([γ-32P]-ATP) (Perkin Elmer)

Sodium dodecyl sulphate (SDS) (Fisher Scientific)

Spectinomycin

Tetramethyl-ethylenediamine (TEMED) (Fisher Scientific)

N-Tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid (TES) (Fisher

Scientific)

Trichloroacetic acid (TCA) (Sigma)

Tris (2-Amino-2-hydroxymethylpropane- 1,3-diol) (Fisher Scientific)

Tryptone soya broth (TSB) (Oxoid)

Uracil (Sigma)

2-Amino-2-hydroxymethyl-propane-1,3-diol (Tris) (Fisher Scientific)

5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal) (Melford

Laboratories Ltd)

5-Fluoroorotic Acid (5-FOA) (Fisher Scientific)

2.1.2. Enzymes

Phusion™ (New England Biolabs)

QuantiTect SYBR green PCR Kit (QIAGEN)

Reverse transcriptase iScript (Bio-Rad)

Taq DNA polymerase (New England Biolabs)

38

Shrimp alkaline phosphatase (Promega)

T4 DNA Ligase (New England Biolabs)

T4 polynucleotide kinase (New England Biolabs)

Restriction endonucleases (New England Biolabs)

RQ RNAse-free DNAse (Promega)

RiboShredder (Cambio)

RNase A (Sigma)

2.1.3. Antibodies

Anti‐FLAG M2 monoclonal antibody, from mouse, Sigma F18041.

2.1.4. Solutions and buffers

10 x TBE buffer: 107.8 g Tris base, 55 g boric acid and 7.44 g EDTA, made up to 1

L with H2O.

Kirby buffer: 1% (w/v) sodium-triisopropylnapthalene sulphonate (TPNS), 6%

(w/v) sodium 4-amino salycilate, 6% (v/v) phenol, made up in 50 mM Tris-HCl,

pH 8.3.

Phenol/chloroform: 25 ml phenol, 24 ml chloroform, 1 ml isoamyl alcohol.

TE Buffer: 10 mM Tris/HCl, pH 8, 1 mM EDTA in 1 L.

Solution I: 50 mM Tris-HCl, pH 8; 10 mM EDTA. Stored at RT.

Solution II: 200 mM NaOH; 1% (w/v) SDS. Stored at RT

Solution III: 3 M potassium acetate, pH 5.5. Stored at 4ºC.

10 x Primer annealing buffer: 100 mM Tris-HCl, pH 8.0, 150 mM NaCl2, 10 mM

EDTA.

STE buffer: 10.3% sucrose, 25 mM Tris HCL pH 8.0, 25 mM EDTA, 2 µg/ µl

lysozyme.

39

Blocking solution: 5ml TBS tween, 2.5 g milk powder, made up to 50 ml with

ultra-pure water.

20 x TBS tween: 500mM Tris, 60mM KCl, 2.8M NaCl, 1.0% Tween 20, pH 7.

IP buffer: 50mM Hepes-KOH at pH7.5, 150mM NaCl, 1mM EDTA, 1% Triton X-

100, 0.1% Na-deoxycholate, 0.1% SDS.

IP Salt buffer: 50mM Hepes-KOH at pH7.5, 500mM NaCl, 1mM EDTA, 1%

Triton X-100, 0.1% Na-deoxycholate, 0.1% SDS.

5X MOPS buffer: 0.2 M MOPS pH 7.0, 0.05 M sodium acetate, 5 mM EDTA. For

1 L of buffer, dissolve in 0.8 L of double autoclaved ultra-pure H2O with stirring,

adjust the pH to 7.0 with 10N NaOH. Bring the final volume to 1L with double

autoclaved ultra-pure H2O water. Dispense into 200ml aliquots and autoclave.

RNA Sample buffer: 10ml deionized formamide, 3.5 ml 37% formaldehyde, 2 ml

5X MOPS Buffer. Dispense into single-use aliquots and store at –20°C in tightly

sealed, screw-cap tubes. These can be stored for up to 6 months.

RNA loading buffer: 50% (v/v) glycerol, 1 mM EDTA, 0.4% (w/v) bromophenol

blue. Prepare in double autoclaved ultra-pure H2O, dispense into single-use aliquots

and store at –20°C.

6 x Orange DNA loading dye: 25 mg Orange G and 30% (v/v) glycerol in 10 ml

and filter-sterilise.

GelRed Stain: 100 ml ultra-pure H2O and 2 µl GelRed (Cambridge Bioscience).

Stored in the dark.

Binding buffer: 20mM Tris-HCl pH7.9, 500mM NaCl, 0.1mM PMSF.

Charge buffer: 50mM NiCl2

Wash buffer: 20mM Tris-HCl pH 7.9, 0.5M NaCl, 60mM imidazole.

Elution buffer: 20mM Tris-HCl pH 7.9, 0.5M NaCl, 0.5M or 1M imidazole.

Strip buffer: 20mM Tris-HCl pH7.9, 0.5M NaCl, 100mM EDTA.

Semi-Dry buffer: 5.82g Tris Base plus 2.92g Glycine plus 0.0375 % SDS and 20 %

Ethanol made in 1 L.

40

2x protein loading dye: 250 mM Tris-HCl pH 6.8, 2% (v/v) SDS, 10% (v/v)

glycerol, 0.01% (w/v) bromophenol Blue and 20 mM DTT.

1x SDS running buffer: 100 ml 10% (v/v) SDS, 30.9 g Tris-HCl, pH 8.3, 144.1 g

glycine in 1 L.

Coomassie Brilliant Blue stain: 0.25% (w/v) Coomassie Brilliant Blue (Sigma),

50% (v/v) methanol, 10% (v/v) acetic acid.

Destainer solution: 20% (v/v) methanol, 10% (v/v) acetic acid.

2.2. Strains, vectors and oligos

2.2.1. Plasmids and expression vectors

Plasmids Features Source/Reference

pN109 E. coli - G. thermoglucosidans conjugative

vector: oriR, oriT, traJ, AmpR, KanR

TMO Renewables

pSX700 E. coli - G. thermoglucosidans conjugative

vector: oriT, traJ, AmpR , KanR

This study

pUCG18 E. coli, G. thermoglucosidans bifunctional

vector: AmpR , KanR

(Taylor, Esteban, & Leak, 2008)

pSET152 Streptomyces conjugative/integrative vector:

lacZα, φC31 int/attP, AprR

(Bierman et al., 1992)

pBlueScript II SK (+) E. coli cloning vector: bla, lacZα, AmpR Stratagene (Alting-Mees & Short, 1989)

pET15b E. coli T7 based His-Tag expression vector:

AmpR

Novagen

pTMO31 G. thermoglucosidans temperature sensitive

suicide vector: pUB110 oriR, KanR

(Cripps et al., 2009)

pG1AK-sfGFP G. thermoglucosidans reporter plasmid: sfGFP,

AmpR , KanR

(Reeve, Martinez-Klimova, de Jonghe,

Leak, & Ellis, 2016)

pG1AK-sfGFP_OriT E. coli - G. thermoglucosidans conjugative

reporter plasmid: sfGFP, AmpR , KanR ,

oriT, traJ,

This study

Table 2.1 list of the plasmids and vectors used during this project.

41

2.2.2. Bacterial strains

E. coli strains

E. coli Strain Description Source/Reference

S17-1 TpRS mR, recA, thi pro hsdR-M+RP4: 2-

Tc:Mu: Km, Tn7 λpir

(Simon, Priefer, & Pühler, 1983)

DH5α F- Φ80 dlacZ, ΔM15 Δ(lacZYA-argF)

U169, recA1, endA1,

hsdR17 (r- m-), supE44 medλ-

thi-1 gyrA, relA1

Invitrogen

BL21 (λDE3) (pLysS) F’ ompT hsdSB (rB- mB-) dcm, gal,

λ(DE3), pLysS, CmR

Novagen

S17-1 Δdcm Δdcm, AprR This study

Table 2.2 List of the E. coli strains used during this project. S17-1 strains were used as donors

during conjugation. DH5α strain was used during routine laboratory cloning and is suitable for

blue/white screening. BL21 (λDE3) (pLysS) strain was used for protein over expression.

G. thermoglucosidans strains

Geobacillus Strain Description Source/Reference

TM444 A derivative of the industrial TM242 Strain (Cripps et al., 2009)

11955 WT isolate David Leak Lab, University of Bath

TM242 Industrial Strain (Cripps et al., 2009)

11955 Δrex Used for analysis of rex This study

TM444 3xFLAG tagged rex Used for analysis of rex This study

Table 2.3 List of the G. thermoglucosidans strains used during this project. Unless otherwise stated

work in this study was carried out using the 11955 strain.

2.2.1. Oligonucleotides

Name/target Sequence (5’->3’)

F_DCMApr ATGCAGGAAAATATATCAGTAACCGATTCATACAGCACCATGTGCAGCTCCATCAGCA

R_DCMApr TTATCGTGAACGTCGGCCATGTTGTGCCTCTTGCTGACGTCCAACGTCATCTCGTTCT

Col_S17-1Dcm_F GCTACCGCAAACCATGCAAA

Col_S17-1Dcm_R TCTCTATCGCCGTATCACGC

UPP-LFlank_Fwd CCTGCAGGTCGACTCTAGAGGCACGGCAGCCCAGTCAA

UPP-LFlank_Rev CCACCCAGGGTCCTCTTTGTTACGGGTTACACTTCC

42

attP_pSET1_Fwd ACAAAGAGGACCCTGGGTGGGTTACAC

attP_pSET1_Rev CTACTTCTTCCGTTGGCGCTACGCTGT

UPP-RFlank_Fwd AGCGCCAACGGAAGAAGTAGAAATTGAAACGCCG

UPP-RFlank_Rev ATTCGAGCTCGGTACCCGGGTCGTTCGCAACATCGCAG

LF-attP_F GAAGATTTGGCGCGCGAA

LF-attP_R CTTGCTGCGCTCGAATTCTC

attP-RF_F GTTACACGACGCCCCTCTAT

attP-RF_R TGTTCCGAAAAGCAAGCTCA

Int_F CCCTCGAGATGACACAAGGGGTTGTG

Int_R GGACTAGTAAAAAAACGCCCCCTTTCGGGGCGCGATTCTACGCCGCTACGTCTTC

Table 2.4 List of the oligonucleotides used to generate and test conjugation in Chapter 3.

Name Sequence (5’->3’) Use/description

F1MuantRex_6.12 ATACGACGAGAATAGCGCCG Δrex generation

R1MuantRex_6.12 ACGTTTTGCGGTTGCTTGTGG Δrex generation

F2MuantRex_6.12 CACCATATTGATTTGGCGGTC Δrex generation

R2MuantRex_6.12 GGAATGTTTTTGCGCCGTGA Δrex generation

Cas2_L TTCAGACAGCTTTGGTCGTG qPCR

Cas2_R CAATCCCTTCAGCGGTAAAA qPCR

Cbd1_L GCGATTTTCCTTGGCATTTA qPCR

Cbd1_R CCCCGTCTTTCAACTCAAAA qPCR

adhE2_L ACAATGCGTACATTCCGTCA qPCR

adhE2_R TATCCGAGCTTGACCATTCC qPCR

Pgi2_F TGGCGGACACACTGTAGAAA qPCR

Pgi2_R ATAAATGGCAAGGCATTGGA qPCR

L_ldh AGCCGGGAGAAACAAGACTG qPCR

R_ldh TGGGTTCGTTGCCACAAGAA qPCR

L_ndh ACGGAAGAAGAAAAGCGGGA qPCR

R_ndh CGGCATAATTGTGGAAGGCG qPCR

L_L-ao TTTTGCATGCGGGAGGAGAT qPCR

R_L-ao TGACGCATTCTCCCTCGATG qPCR

L_hcbd CCGGCGCTAAGGAAGAGAAA qPCR

R_hcbd CTCGGAGTGCGGATCAAGTT qPCR

L_mhk CTGCTGGCAAGAGGGAACAG qPCR

R_mhk CTCCGCTTCAGACTCCATCA qPCR

L_Fnr GGATGGACGCAGCAGAACTA qPCR

R_Fnr ATTTCTCCAGGGCCGCAAAT qPCR

L_mABC GCCGCAAGTGTCGTTTCTTT qPCR

R_mABC TTCATTTGCTCCATCGACGC qPCR

Table 2.5 List of the oligonucleotides used to generate and test a rex mutant.

43

Name Sequence (5’->3’)

GRex_reg_F CGGATCCTATGATATGTGTAAAGCGCTGG

GRec_reg_R CGGATCCACATCACAACAACCGGCAGCTG

GRex_Flag_F CACGATATCGACTACAAAGACGATGACGATAAGTAAAGGAGAGAACCAAAAGTG

GRex_Flag_R GTCCTTATAATCGCCGTCATGATCTTTGTAGTCTTCGGCTGGATAATTTTTCAA

L_cplpny_Re-FLG ATATTGATTTGGCGGTCGAG

R_cplpny_Re-FLG GCCCTTTTGTCGCATCTTTA

Table 2.6 List of the oligonucleotides used to generate and test a 3xFLAG tagged rex strain.


CT1T_F CCTAACTGTTGTTAAATTGTTAACATCCTAACTGC

CT1T_R GCAGTTAGGATGTTAACAATTTAACAACAGTTAGG

CT1G_F CCTAACTGTTGTGAAATTGTTCACATCCTAACTGC

CT1G_R GCAGTTAGGATGTGAACAATTTCACAACAGTTAGG

GTLD_F CCTAACTGTTGTGAAATAATGCACAACCTAACTGC

GTLD_R GCAGTTAGGTTGTGCATTATTTCACAACAGTTAGG

C04422_F ATTTATTTTTTGTGAAAACTATATAAAACTTTCAT

C04422_R AATGAAAGTTTTATATAGTTTTCACAAAAATAAAT

Table 2.7 List of the oligonucleotides used during EMSA experiments.


autrex/RBSgfp_F GCACTCTCCTCTTTGTTACGG

autrex/RBSgfp_R CCAGAATAGGGACGACACCA

GRex_S_top TCGAATTGAATGATACCGATGTTTGTGATAGAATTGCTTGTGAAATAATGCACAAT

GRex_S_bottom CTAGATTGTGCATTATTTCACAAGCAATTCTATCACAAACATCGGTATCATTCAAT

GRex_D_top TCGAATTGAATGATACCGATGTTTGTGATAGAATTGCTTGTGAATGTGAACACATTCACAAT

GRex_D_bottom CTAGATTGTGAATGTGTTCACATTCACAAGCAATTCTATCACAAACATCGGTATCATTCAAT

CRex_S_top TCGAATTGAATGATACCGATGTTTGTGATAGAATTGCTTGTTAAATTGTTAACATT

CRex_S_bottom CTAGAATGTTAACAATTTAACAAGCAATTCTATCACAAACATCGGTATCATTCAAT

CRex_D_top TCGAATTGAATGATACCGATGTTTGTGATAGAATTGCTTGTTAATGTTAAAACATTAACAAT

CRex_D_bottom CTAGATTGTTAATGTTTTAACATTAACAAGCAATTCTATCACAAACATCGGTATCATTCAAT

qPCR_1Rex_Fwd TTCTTATTTCGGTGCGCTTGG

qPCR_1Rex_Rev CCAAGTTTCCGACGCCAAAT

CRexqPCR_F TATTCGTCGTCTGCCTCGTT

CRexqPCR_R GCTTGCGGTAATGCCCATAC

GRexqPCR_F CAATGAGCAGCCAAAGATCCC

GRexqPCR_R CACGCTGTTTTCCTGAAGCA

Table 2.8 List of the oligonucleotides used to construct and test G/C-Rex promoter/operator

expression circuit.

44

2.3. Growth media, storage and selection

2.3.1. Media

E.coli Media Use and preparation Ingredients per litre of media

Lennox broth (LB) Liquid media used for routine growth of E. coli

strains. 50 ml aliquots in universal bottles were

sterilised by autoclaving. LB media was prepared

with no glucose for BACTH assays.

10 g Difco Bacto-tryptone, 5 g Difco

yeast extract, 5 g NaCl and 1 g glucose.

Lennox agar (LA) Prepared as LB, but with the addition of 1.5 g of

agar per 100 ml. Flask was stoppered with foam

bung, covered with foil and autoclaved.

15g agar, 10 g Difco Bacto-tryptone, 5

g Difco yeast extract, 5 g NaCl and 1 g

glucose.

Low salt LB/LA Similar to LA but used when selecting for KanR 15g agar, 10 g Difco Bacto-tryptone, 5

g Difco yeast extract, 1 g glucose.

2xYT A rich liquid medium used for high efficiency

transformation. 10 ml aliquots in universal bottles

were sterilised by autoclaving.

16 g Difco Bacto-tryptone, 10 g Difco

Yeast Extract, 5 g NaCl.

SOC Rich media for use after high efficiency

transformation. Adjust pH to 7.0 with 10N NaOH.

Made up to 1 L with distilled water. Autoclave to

sterilize.

20 g tryptone, 5 g yeast extract, 10 mM

NaCl, 2.5 mM KCl, 10 mM MgCl2, 10

mM MgSO4, 20 mM glucose.

Table 2.9 List of the media used for E. coli growth.

G. thermoglucosidans

Media

Use and preparation Ingredients per litre of media

Tryptone soya broth

(TSB)

General growth media. Made up to 1 L in distilled

water. Autoclave to sterilize.

30g tryptone soya broth powder

Tryptone soya agar

(TSA)

General growth media. Made up to 1 L in distilled

water, 100ml aliquoted into conical flasks and then

1.5 g agar is added before autoclaving.

30g tryptone soya broth powder, 15g

agar

USM Fermentation tube culture evaluations of strains uses

the USM media base. When making USM seed

media, the ingredients except for biotin are made up

to ~500ml then pH adjusted to 6.0. The biotin, yeast

extract and C-source are then added before pH

adjusting to 6.7 and the final volume is made up to 1

L. In standard seed medium, yeast extract is added at

a concentration of 20 g/L and glycerol at 30 g/L

Store out of direct light or in an opaque bottle.

12.4 ml 2M NaH2PO4, 12.5 ml 4M Urea,

50 ml 0.5M K2SO4, 5 ml 1M Citric acid,

12.5 ml 0.25M MgSO4, 2.5 ml 0.02M

CaCl2, 12.5 ml Sulphate TE solution, 94

µl 0.8 mg/ml Biotin, 0.25 ml 10 mM

Na2MoO4

Sulphate TE solution Made up to 1 L in distilled water. Store out of direct

light or in an opaque bottle in fridge.

Conc. H2SO4 5 ml, ZnSO4.7H2O 1.44g,

FeSO4.7H2O 5.56g, MnSO4.H2O 1.69g,

CuSO4.5H2O 0.25g, CoSO4.7H2O

0.562g, H3BO3 0.06g, NiSO4.6H2O

0.886g

45

Minimal media

(MM)

Solutions were combined after each had been

autoclaved separately. Made up to 1 L in distilled

water.

20% (v/v) M9 salt (5x) solution, 80%

(v/v) casamino acids + D-glucose

solution

Casamino acids and

D-glucose solution

Made up to 1 L with RO water and the solution was

autoclaved.

1.25g casamino acids and 12.5g D-

glucose

M9 salt (5x) solution The pH was adjusted to pH 7.5 and the solution

autoclaved.

1.5g K2SO4, 1.03g Na2HPO4, 5g NH4Cl,

3.7g MgSO4.7H2O, 0.015g MnCl2.4H20,

0.025g CaCl2.H2O, 0.021g FeCl3, 10mM

Tris-HCl

R2YE trace elements Made up to 1 L with RO water. Store out of direct

light or in an opaque bottle in fridge.

40mg ZnCl2, 200mg FeCl3.6H2O

10mg CuCl2.2H2O, 10mg MnCl4.4H2O,

10mg Na2B4O7.10H2O, 10mg

(NH4)6Mo7O2.4H2O

Table 2.10 List of the media used for G. thermoglucosidans growth.

2.3.2. Antibiotics and additives

Name Stock concentration Stock solvent Working concentration

(liquid media)

Working

concentration (solid

media)

Ampicillin 100 mg/ml 60% ethanol 100 μg/ml 100 μg/ml

Apramycin 50 mg/ml dH2O (Filter-

sterilised)

20-50 μg/ml 20 μg/ml

Choramphenicol 34 mg/ml 100% ethanol 34 μg/ml 25 μg/ml

IPTG 1 M dH2O (Filter-

sterilised)

1 mM 0.5 mM

Kanamycin 50 mg/ml dH2O (Filter-

sterilized)

12.5 μg/ml 12.5 μg/ml

X-gal 40 mg/ml N, N-dimethyl-

formamide

N/A 40 μg/ml

5-FOA 100 µg/ml DMSO (Filter-

sterilised)

1-200 µg/ml 1-200 µg/ml

Uracil 10 µg/ml 100% ethanol 1-10 µg/ml 1-10 µg/ml

Table 2.11 List of antibiotics and additives used in growth media for selection.

46

2.3.3. Growth and storage of bacterial strains

Growth of E. coli cultures and storage

E. coli strains were grown at 37°C on solid media or in liquid media shaking at 250

rpm. Colonies formed on solid media plates were kept for up to 3 weeks at 4°C. To

achieve long-term storage, overnight LB cultures were suspended in a final

concentration of 20% (v/v) glycerol, and stored at -80°C.

Preparation of chemically competent E. coli by CaCl2

method

Chemically competent E. coli cells were prepared by inoculating 5 ml of LB media

from glycerol stocks and grown overnight at 37 ºC with shaking at 250 rpm. The

overnight culture was used to inoculate (by a ratio of 1:100) 50 ml of LB media in a

250 ml conical flask and incubated at 37 ºC with shaking at 250 rpm until it reached

an OD600 of 0.4. The culture was harvested by centrifugation at 4ºC, 2000 x g for 8

min. The cell pellet was resuspended in 20 ml of chilled 50 mM CaCl2, incubated in

an ice bath for 20 min, and centrifuged again at 4ºC, 2000 x g for 8 min. The pellet

was gently resuspended in 2.5 ml of chilled 0.1 M CaCl2, 10% glycerol. 100 µl

aliquots were frozen in a dry ice ethanol bath and stored at -80°C.

Preparation of electrocompetent E. coli

Electrocompetent E. coli cells were prepared by inoculating 5 ml of LB media from

glycerol stocks and grown overnight at 37 ºC with shaking at 250 rpm. The

overnight culture was used to inoculate (by a ratio of 1:100) 50 ml of LB media in a

250 ml conical flask and incubated at 37 ºC with shaking at 250 rpm until it reached

an OD600 of 0.35. The culture was harvested by centrifugation at 4ºC, 2000 x g for

10 min. The cell pellet was resuspended in 50 ml of chilled 10% v/v glycerol. This

was centrifuged again at 4ºC, 2000 x g for 5 min and resuspended in 25 ml of

chilled 10% v/v glycerol. A further centrifugation at 4ºC, 2000 x g for 10 min

followed and the pellet resuspended in 10 ml of chilled 10% v/v glycerol. A final

centrifugation followed at 4ºC, 2000 x g for 10 min and the pellet resuspended in

the residual volume ~1.5ml. 50 µl aliquots were frozen in a dry ice ethanol bath and

stored at -80°C.

47

Growth of G. thermoglucosidans cultures and storage

G. thermoglucosidans strains were grown at 60°C on solid media or in liquid media

shaking at 250 rpm, all media and equipment was pre-warmed to reduce

temperature stress. To grow G. thermoglucosidans, frozen stocks were used to

inoculate solid media (e.g. TSA) and incubated overnight. Plates were inoculated

by spreading 150 µl of glycerol stock or a sterile loopful of biomass onto a plate.

Distinct colonies or confluent growth was harvested and the biomass was dispersed

in a small amount of liquid media before being used to inoculate further pre-

warmed media.

Aerobic growth of a 50 ml culture was achieved in a 250 ml flask. Oxygen limited

growth was achieved by completely filling a 15 ml tube or a 1.5 ml microfuge tube

with culture and tightly sealing it. When grown on solid media, agar plates were

placed in plastic bags including a damp cloth to prevent drying out. Colonies that

formed on solid media plates were kept for 1 week at 4°C before being discarded.

To achieve long-term storage, overnight cultures were suspended in a final

concentration of 33% (v/v) glycerol, and stored at -80°C.

Preparation of electrocompetent G. thermoglucosidans

cells

Electrocompetent G. thermoglucosidans cells were prepared by inoculating 50 ml of

TSB media from glycerol stocks and grown overnight at 52ºC with shaking at 250

rpm overnight. 5 ml of the overnight culture was used to inoculate 50 ml of TSB

media in a 250 ml conical flask and incubated at 55ºC with shaking at 250 rpm until

it reached an OD600 of between 1.5 - 2.0. The culture was placed on an ice bath for

10 min and cells harvested by centrifugation at 4ºC, 2000 x g for 15 min. The cell

pellet was resuspended in 50 ml of chilled electroporation buffer (0.5 M sorbitol, 0.5

M mannitol and 10% v/v Glycerol). The cell pellet was sequentially washed in 30

ml, 20 ml, 10 ml of chilled electroporation buffer. A final centrifugation at 4ºC,

2000 x g for 15 min followed and the pellet resuspended in the residual volume of

~2 ml. 60 µl aliquots were frozen in a dry ice ethanol bath and stored at -80°C.

48

2.4. DNA and RNA manipulation and molecular cloning

2.4.1. DNA gel electrophoresis

To analyse DNA size and quality agarose gel electrophoresis was used. DNA

samples were prepared by adding 2µl Orange G loading dye, DNA sample 1-10 µl,

H2O make up to a total of 12 µl before loading into the gel wells. Gels of between

0.8-1% agarose, in TBE buffer were made and run at 2 V/cm. Gels were stained

after electrophoresis using GelRed Nucleic Acid Gel Stain and visualised using a

UV transilluminator.

2.4.2. RNA gel electrophoresis

To analyse RNA size and quality it must be separated on a denaturing gel to disrupt

secondary structure. RNA samples were prepared by adding RNA Sample buffer in

a ratio of 1:2 to make a total volume of 10 µl. Samples were then heated to 60°C for

5 min and allowed to cool for 2 min before 2 µl of RNA Loading buffer was added.

Gels consisted of 36.5 mM MOPS, 9.1 mM sodium acetate, 0.9 mM EDTA, 2 M

formaldehyde, and 1.2% agarose and were run at 4-5 V/cm. Gels were stained

using GelRed Nucleic Acid Gel Stain and visualised using a UV transilluminator.

2.4.3. DNA gel purification

For extraction and purification DNA from agarose gel QIAquick™ or MinElute™

Gel Extraction Kits were used according to manufacturer’s instructions.

2.4.4. General polymerase chain reaction (PCR)

Every reaction consisted of 1 μl 50 ng/μl template DNA, 5 μl 10 pmol/μl of each

primer, 5 μl 10x Polymerase buffer, 1.5 μl 10 mM dNTPs, 1 μl polymerase. Each

reaction was made up to 50 μl with ultra-pure H20. Reactions were also

supplemented with either 2.5μl DMSO if appropriate. The polymerase used was

determined by the intended purpose of the product: Phusion if product is to be used

downstream, Taq if diagnostic.

49

Thermocycler conditions:

1. Initial denaturation: 2 min at 95 °C

2. Denaturation: 30 sec at 95 °C

3. Annealing: 30 sec at 50-68 °C

4. Extension: ~30 sec/kb of product at 72 °C

5. Repeat steps 2-4 for 30 cycles

6. Final extension: 5 min at 72°C

7. Hold at 4 °C

The conditions of the annealing and extension phases were optimised for each

reaction according to the primer and template used.

2.4.5. Inverse PCR

The reaction was prepared as described in Section 1.4.4, however the primers were

phosphorylated and for larger products such as vectors (>3kb), the following PCR

cycle was used:

1. Initial denaturation: 2 min at 95 °C

2. Denaturation at 95 °C for 1 min

3. Annealing at 55-65 °C for 45 s

4. Extension at 72 °C for 2.5 min per 1 kb


6. Denaturation at 96 °C for 1 min

7. Annealing at 55-65 °C for 45 s

8. Extension at 72 °C for 3.5 min per 1 kb


10. Final Extension at 72 °C for 15 min

11. Hold at 4 °C

2.4.6. Colony PCR

Colony PCR was used to identify mutants or genetically altered strains. A single

colony was picked from a plate and biomass suspended in sterile ultra-pure H2O

50

and heated to 100 °C for 10 min. Cell debris was removed by centrifugation and 5

μl of the supernatant used as template DNA in a standard Taq PCR.

2.4.7. DNA ligation

Ligation reactions consisted of 1 μl 10X T4 DNA Ligase buffer, 1 μl T4 DNA

ligase, 100 ng vector DNA, the appropriate quantity of insert DNA (usually a ratio

of 1:3 see equation). This was made up to a total volume of 10 μl. The reaction

proceeded at either room temperature for 2 h, 4 °C for 16 h, or 16 °C for 4 h.

2.4.8. Oligonucleotide annealing for construction of dsDNA fragments

Equal molar amounts of each primer mixed with 10 x Annealing Primer buffer

(diluted to 1 x in reaction). This was made up to volume with ultra-pure H2O,

heated to >95 ºC for 2 min, then cooled to 50 ºC over at least 30 min. Placed on ice

until use or stored at -20 ºC.

2.4.9. Restriction digest

Restriction digests contained ~1 μg DNA, ~10 U of each enzyme, 2 μl 10X

Restriction buffer and were made up to a total volume of 20μl using ultra-pure H2O.

Digestion occurs over 2-4 hs at the temperature recommended by the manufacturer

of the enzyme. Samples were analysed via agarose gel electrophoresis for the

100 ng X size of insert (kb) Insert (ng) = X ratio (e.g 1/3)

Size of vector (kb)

Figure 2.1 Equation used to calculate amount of DNA required for an optimal ligation reaction.

51

production of the desired fragment sizes. Fragments, if necessary for downstream

use, were isolated from the agarose gel and purified.

2.4.10. Dephosphorylation of DNA

To inhibit self-ligation 5’ ends of digested vectors were dephosphorylated. 1 μl

shrimp alkaline phosphatase, ~1 μg restriction digested vector, 3.5 μl 10 x

Phosphatase buffer were mixed in a final volume of 35 μl with ultra-pure H2O and

incubated at 37 °C for 30 min, then heat inactivated at 65 °C for 20 min.

2.4.11. Phosphorylation of DNA primers

4 μl 200 pmol/μl primers were phosphorylated using 1 μl T4 polynucleotide kinase,

2 μl 10X Kinase buffer, 2 μl 10 mM ATP. Made up to 20μl with ultra-pure H2O and

incubated at 37 °C for 20 min, and then heated at 90 °C for 2 min to inactivate the

enzyme before using for PCR.

2.4.12. Determination of DNA/RNA concentration

Determination of RNA or DNA concentrations was achieved using a NanoDrop

1000 (Thermo Fisher Scientific). Between 0.5-2 μl of sample was loaded and

measured to give an approximation of the concentration. For a more accurate

measurement the Qubit High Sensitivity DNA assay kit was used along with the

Qubit 2.0 Fluorometer.

2.4.13. Gibson assembly

PCR amplified fragments with overlapping regions (20 bp) were Gibson assembled

in one reaction. 5 μl of the DNA fragments (0.5 pmol of each fragment) were added

to 15 μl of Gibson Assembly master mix (NEB) and incubated at 50 oC for 1 h. 1-2

μl was then used in a transformation with extra competent E. coli (High Efficiency,

NEB #C2987) cells.

52

2.4.14. DNA sequencing

Sequencing of DNA was carried out externally by Eurofins Genomics using the

value read service. 15 μl of 50-100 ng/μl purified plasmid samples were sent for

each strand to be sequenced in 1.5 ml Eppendorf tubes.

2.4.15. Generating an in-frame disruption strain using λ RED

This technique can be used for gene targeted deletion and uses a λ Red recombinase

under the control of an inducible promoter on a low copy plasmid (Gust, Rourke,

Bird, Kieser, & Chater, 2004). It was applied in this project to generate a S17-1

Δdcm strain of E. coli. To target the gene of interest (F_DCMApr and R_DCMApr)

primers were used that amplify an apramycin antibiotic resistance cassette. These

primers had homology of 39 nt in length to the S17-1 dcm gene that was to be

inactivated. The resulting PCR product therefore contained an apramycin resistant

gene flanked by regions of dcm homology. S17-1 E. coli cells were transformed with

the λ-Red recombination plasmid pIJ970 and Electocompetent S17-1 (pIJ970) cells

were then transformed with 100 ng of the dcm::apr allele cassette (PCR product).

Transformants were grown over night at 30 oC on LA plates containing

chloramphenicol to select for pIJ970, 1 mM L-arabinose to induce λ-Red mediated

hyper-recombination, and apramycin to select for recombination of cassette. Ex-

transformants were then grown at a raised temperature of 37 oC to induce the loss

of the pIJ970 plasmid. The strain was screened for the loss of CmR and gain of

AprR, then analysed by colony PCR and methylation specific restriction digest.

2.4.16. DNA purification

Small scale miniprep plasmid purification

A cell-pellet resulting from a ~5 ml culture was resuspended in 200 μl of solution I.

Immediately 400 μl of solution II was added and mixed by inverting. 300 μl of

solution III was then added and the contaminating RNA was degraded by the

addition of 1 μl 10 mg/ml RNAse A and incubated for 10 min at room

temperature. The samples were centrifuged at 16,100 x g for 5 min and the

supernatant extracted with 150 μl phenol/chloroform. The mixtures were vortexed

53

for 2 min, centrifuged as before and precipitated with 600 μl isopropanol. The

samples were incubated on ice for 10 min and centrifuged as above. The pellet was

washed with 70% ethanol, allowed to dry and finally resuspended in 30 μl TE

buffer or ultra-pure H2O.

Large scale midiprep plasmid purification

Each preparation required a cell pellet from 50 ml of culture and QIAGEN Plasmid

Midiprep kit was utilised for plasmid purification and the manufacturer’s procedure

followed.

Chromosomal DNA extraction

For G. thermoglucosidans chromosomal DNA extraction and purification 5 ml of

culture was harvested by centrifugation at 13,000 x g for 1 min and the supernatant

discarded. The pellets were washed with 1 ml 10.3% sucrose and then resuspended

in 250 µl of STE buffer and incubated at 37 °C for 30 min. 330 µl of Kirby mix was

added and vortexed for 30 sec. 670 µl of phenol/ chloroform was added and

vortexed for another 30 sec before centrifuging again at 13,000 x g for 5 min. The

upper phase was transferred to a new tube and 250 µl of phenol/ chloroform was

added, vortexed, and centrifuged at 13,000 x g 4 °C for 5 min. The upper phase was

transferred to a new tube and an equal volume of isopropanol and a 1/10 volume of

3 M sodium acetate was added and mixed by inversion. The mixture was stored at -

20 °C for 1 h. This was again centrifuged at 13,000 x g 4 °C for 5 min and the

supernatant carefully discarded. The pellet was resuspended in TE buffer and

incubated for 30 min at 37 °C with 10 µl/ml of RNase A. Chromosomal DNA was

extracted with 100 µl of phenol/chloroform and precipitated with equal volume of

isopropanol and a 1/10 volume of 3 M sodium acetate as before. The pellet was

washed with 150 µl of ethanol and left to air dry for 5 min before being resuspended

in 100 µl of TE buffer.

54

2.4.17. RNA Isolation and purification

Cryogenic grinding method

To isolate RNA from G. thermoglucosidans, 15 ml of cell culture was added to 400 µl

of ice-cold 95% ethanol, 5% phenol/chloroform/isoamyl alcohol (25:24:1) mix per

ml of culture to protect RNA, and then vortex mixed. Samples were then

centrifuged at 6,000 x g 4 °C for 10 min and the supernatant carefully discarded.

The pellet was vortexed in 1ml of ice cold TE buffer to dislodge it and small

droplets flash frozen in liquid nitrogen. The cells were then placed into cryogenic

grinding apparatus and beaten for 1.5 min at 30 Hz for three cycles, cooling in

liquid nitrogen between cycles. After removing the lysed cells from the cryogenic

grinding apparatus a Qiagen RNeasy midi kit was utilised to purify RNA according

to the manufacturer’s instructions.

Rapid high throughput sonication method

To rapidly isolate RNA from G. thermoglucosidans 1.5ml of culture was harvested by

centrifugation at top speed at 4 °C for 30 sec. The pellet was immediately

resuspended in chilled 400 µl Kirby mix. Lysed cells were sonicated using a

Diagenode Bioruptor sonicating waterbath for 10 cycles of 30s on/30s off pulses.

To isolate RNA from the cell debris 300 µl of phenol/chloroform was added,

vortexed for 1 min, and centrifuged at 13,200 x g 4 °C for 5 min and the upper

phase transferred to a new tube. 200 µl of phenol/ chloroform was added, vortexed,

and centrifuged at 13,200 x g 4 °C for 5 min. The upper phase was transferred to a

new tube and an equal volume of isopropanol and a 1/10 volume of 3 M sodium

acetate was added and mixed by inversion. The mixture was stored at -20 °C for 1

h. This was again centrifuged at 13,200 x g 4 °C for 5 min and the supernatant

carefully discarded. The pellet was washed with 150 µl of ethanol. The pellet was

resuspended in 200 µl of 1x DNAse buffer and 0.5 µl of RNAse-free DNAse added

and incubated at 37 °C for 30 min. After incubation 200 µl of RNAse-free H2O was

added and 200 µl of phenol/chloroform, then vortex for 1 min. Centrifuged at

13,200 x g 4 °C for 5 min and the upper phase was transferred to a new tube and an

equal volume of isopropanol and a 1/10 volume of 3 M sodium acetate was added

and mixed by inversion. The mixture was stored at -20 °C for 1 h and centrifuged at

55

13,200 x g 4 °C for 5 min and the supernatant carefully discarded. The pellet was

washed with 150 µl of ethanol and left to air dry for 5 min before the pellet was

resuspended in 50 µl of RNAse-free H2O. Samples were analysed for quality by gel

electrophoresis (28S RNA and 18S RNA band intensity) and NanoDrop and stored

at -80 °C.

2.4.18. DNA introduction into bacteria

E. coli transformation

Heat shock

A 150 µl aliquot of chemical competent cells was thawed on ice, mixed with 1-2 µl

of DNA (10-300ng) and gently mixed by pipetting, then incubated on ice for 10-30

mins. The vial was then heat shocked at 42 °C for exactly 30 sec and immediately

placed back on ice for 5 min. 950 µl of pre warmed SOC or LB was added and the

vial incubated at 37 °C shaking at 250 rpm for 1 h. 50-100 µl of the transformation

mix was then plated onto pre warmed LA plates containing the appropriate

antibiotic. For low efficiency transformations the transformation mix was

centrifuged at 13,000 x g for 30 sec, the pellet resuspended in a residual ~50 µl and

then plated out as outlined before.

Electroporation

A 60 μl aliquot of electrocompetent cells was thawed on ice, mixed with 1-2 µl of

plasmid DNA (~100 ng) and gently mixed by pipetting, then incubated on ice for

10 min. The mixture was then transferred to an ice-cold (1 mm gap) electroporation

cuvette and electroporated at 2500 V (capacitance 10 µF, and resistance 600 Ω, time

constant 4-6). Immediately after pulsing, 500 μl of pre warmed SOC or LB media

was added to the cuvette and the suspension transferred to an Eppendorf tube

containing an additional 500 μl of pre warmed media. The cells were incubated at

37 °C for 1 h with shaking at 250 rpm and plated on pre-warmed LA plates

containing appropriate selective antibiotic.

56

G. thermoglucosidans transformation

Electroporation

A 60 μl aliquot of electrocompetent cells was thawed on ice, mixed with 1-2 µl of

plasmid DNA (~100ng) and gently mixed by pipetting, then incubated on ice for 10

mins. The mixture was then transferred to an ice-cold (1 mm gap) electroporation

cuvette and electroporated at 2500 V (capacitance 10 µF, and resistance 600Ω, time

constant 4-6). Immediately after pulsing, 500 μl of pre warmed TSB media was

added to the cuvette and the suspension transferred to an Eppendorf tube

containing an additional 500 μl of pre warmed TSB. The cells were incubated at

55 °C for 2 h with shaking at 250 rpm and plated on pre-warmed TSA plates

containing appropriate selective antibiotic. Colonies usually became visible after 1

day incubation, but could take longer to grow.

Conjugation

For conjugation into G. thermoglucosidans, E. coli strain S17-1 Δdcm was used. 100 μl

of E. coli overnight culture was used to inoculate 5 ml LB containing 100 μg/ml

ampicillin to select for the conjugative vector. This was grown at 37 °C with

shaking at 250 rpm to an OD600 0.4-0.6. The cells were harvested by centrifugation

at 6,000 x g at 37 °C for 6 min and washed twice in pre warmed LB (37 °C). G.

thermoglucosidans was grown by using biomass to inoculate pre warmed (60 °C) TSB

and grown till OD 600 0.4 at 60 °C with 250 rpm shaking. This was then cooled to

37 °C before being added in a 1:9 ratio to washed and pelleted E. coli cells and

vortex mixed to resuspend all cells. The mixed G. thermoglucosidans and E. coli cells

were then centrifuged at 6,000 x g at 37 °C for 10 min and resuspended in residual

~100 μl of pre warmed (37 °C) TSB per 10 ml of mixed cells. 100 μl of cells were

plated onto pre-warmed (37 °C) 20 ml TSA + 10 mM MgCl2 plates, allowed to dry

in 37 oC for 5 min before placing lids. Plates were incubated at 37 oC for between 1-

3 hs to allow conjugation to occur. Plates were then incubated at 60 oC for 1-3 h to

allow G. thermoglucosidans to recover. To select for uptake of the conjugative vector,

an antibiotic overlay was applied evenly. Each plate was covered in a dilution of

kanamycin (5 μl of Kan 50 mg/ml, in 500 µl of TSB) and allowed to dry in 60 oC

incubator for 10 min before replacing lids. The final concentration of selection was

57

12.5 μg/ml kanamycin. Plates were incubated overnight to produce visible ex-

conjugate colonies.

2.4.19. Analysis of nucleic acids

qRT PCR

RNA samples were reverse transcribed using High Capacity RNA-to-cDNA Kit

(Thermo Fisher Scientific) according to the manufacturer’s instructions. RT-qPCR

experiments were performed using a 96-well plate and a StopOnePlus Real-Time

PCR system (Applied Biosystems). Each reaction contained 12.5 μl maxima SYBR

Green qPCR master mix, 2 μM of a specific set of up and downstream primers, 1 μl

of template cDNA (cDNA was diluted 1/50) and made up to a total of 25 μl with

nuclease free H2O. Data was analysed by StopOnePlus Applied Biosystems Real-

Time PCR Software.

2.5. Protein expression and purification

2.5.1. Protein expression

BL21 (pLysS) transformants were inoculated into LB with 100 μg/ml ampicillin

and 34 μg/ml chloramphenicol and incubated overnight at 37°C with 250 rpm

shaking. The overnight culture was used to inoculate 20-250 ml cultures grown to

an OD600 of 0.4-0.7. The cultures were cold-shocked in an ice bath for 15 min to

induce cold-shock proteins. Protein expression was then induced by adding 1 mM

IPTG and grown at 30-37°C for 3 h with 250 rpm shaking.

2.5.2. Protein purification

Cell pellets were resuspended in Binding buffer, 25 μg/ml PMSF and a protease

inhibitor tablet (Roche) was added. This was sonicated 6 x 10 s with 50 s intervals

at 30% amplitude. The cell suspension was centrifuged at 20,000 x g for 10 min at

4 °C. The cleared supernatant was then loaded on a Ni-affinity column for

purification.

58

2.5.3. Nickel sepharose hand-made column

A Ni-NTA sepharose column was prepared by filling the tip of a syringe with glass

wool and then topping with ~4ml of iminodiacetic acid (IDA) sepharose fast-flow

resin. The column was washed with 3 column volumes (CV) of dH2O and Ni-

charged with 5 CV of Charge buffer. The column was washed again with Binding

buffer and the cleared supernatant loaded onto the column. The column was

washed with 10 CV of Binding buffer 5 CV of Wash buffer. 1-2 CV of Elution

buffer was applied and the samples collected from the flow-through and analysed

by SDS polyacrylamide gel. Before storage at 4oC the column was washed with 6

CV of Strip buffer and 3 CV of dH2O then stored in 20% ethanol.

2.5.4. Concentrating Protein

Protein samples were concentrated using VIVASPIN-6 membrane 5,000 MWCO

PES (Sartorius stedim biotech) and VIVASPIN-500 membrane 3,000 MWCO PES

(Sartorius stedim biotech) by following the manufacturer’s instructions. To prepare

the membrane dH2O was added to the column and centrifuged at 2,000 x g for 5

min at 4 °C. The protein sample was centrifuged at 10,000 x g at 4 °C until the

required concentration or volume was achieved.

2.5.5. Determining a protein concentration

Bradford assay

Protein quantification using the Bradford assay is based on an absorption shift of

Brilliant Blue G dye in the Bradford Reagent (Sigma) from 465 to 595 nm when the

protein binds. To measure the concentration of an unknown protein concentration

1-200 μl of protein was added to 300-499 μl dH2O and 500 μl Bradford reagent.

The absorbance was measured and compared to a standard curve previously

generated using a variety of known concentrations of Bovine Serum Albumin

(BSA).

Qubit

Protein samples were accurately determined using a Qubit protein assay and Qubit

2.0 Fluorometer according to manufacturer’s instructions.

59

2.6. SDS polyacrylamide gel

SDS polyacrylamide gels were prepared using 15% resolving gel and a 5% stacking

gel. The equipment used was a Hoefer Mighty Small dual gel caster (Hoefer miniSE

260 unit– Amersham Biosciences) and was assembled as per manufacturer’s

recommendation. The resolving gel consisted of 5 ml 30% (v/v) acrylamide/bis-

acrylamide mix solution (Severn Biotech), 2.5 ml 1.5 M Tris pH 8.8, 0.1 ml 10%

(v/v) SDS, 50 μl 10% (w/v) Ammonium Persulphate (APS), 20 μl TEMED and

2.33 ml ultra-pure H2O. This was poured to fill three quarters and allowed to set

before the stacking gel was added. 70% ethanol was used to cover the gel as it set

and was then rinsed away with ultra-pure H2O. The stacking gel consisted of 0.75

ml 30% (v/v) acrylamide/bis-acrylamide mix solution (Severn Biotech), 1.25 ml 0.5

M Tris pH 6.8, 50 μl 10% (v/v) SDS, 50 μl 10% APS, 8 μl TEMED and 2.89 ml

ultra-pure H2O. A comb was inserted carefully ensuring no air bubble was

introduced immediately after the stacking gel was poured. The gel was left to set for

30 min before use.

2.6.1. Sample preparation

Protein samples were prepared to load on the gel by adding 2 x protein loading dye

and DTT. The samples were denatured at 100 °C for 5 min and then run on the

SDS polyacrylamide gel on Hoefer SE260, a mini-vertical gel electrophoresis unit

(Amersham Biosciences) at 200 V for 70 min using 1x SDS Running buffer. The

protein bands were visualized by staining with Coomassie Brilliant Blue and the

background removed by a destainer solution.

2.7. Western blot

2.7.1. Semi-dry transfer

Samples of 3xFLAG tagged protein were analysed via western blot. Samples were

normalised (using a Bradford assay or Qubit protein assay) and run on an SDS gel.

Six sheets of blotting paper were cut to the desired size and soaked in Semi-dry

buffer. The SDS gel was placed down on three sheets of soaked blotting paper then

a single sheet of nitrocellulose membrane (Thermo Fisher) was then placed on top

60

of the gel followed by a further three sheets of soaked blotting paper. The transfer

was run at 20 V, 150 mA for 1 h in a Hoefer semiPhor tank before undergoing

washes and antibody attachment.

2.7.2. Washes and antibodies

The nitrocellulose membrane with the transferred proteins was incubated in

blocking solution for 30 min at room temperature with slow agitation. The

membrane was washed, incubating with 2x TBS tween for 10 min at room

temperature with slow agitation. It was then incubated overnight at 4 oC in 1/1000

dilution of primary antibody, which was diluted in freshly made blocking solution,

with slow agitation. The membrane was again washed in 2x TBS tween twice more

at 4 oC. The membrane was then incubated in an appropriate dilution of secondary

antibody HRP conjugate, made in fresh blocking solution at 4 oC for 1 h. The

membrane was again washed 3 times with 2x TBS tween for 10 min and then

subjected to ECL detection.

2.7.3. ECL detection and development

The Amersham™ ECL™ Prime Western Blotting Detection Reagent (GE

Healthcare) was used according to the manufacturer’s instructions. 1 ml detection

reagent and 1 ml of enhancer solution (GE Healthcare) were mixed together, added

to the membrane and the detection reagent added. The mixture was carefully

pipetted over the membrane for 1 min. The membrane was then removed with

forceps and gently touched on the filter paper to drain excess solution. The soaked

membrane was then wrapped in Saran wrap and exposed to film for between 20 sec

– 5 min depending on the intensity of the bands.

2.8. Chromatin immunoprecipitation (Chip –seq)

The DNA was crosslinked with the protein by adding 37% formaldehyde to the

culture making a final concentration of 1%. Incubation was continued as before for

20 min, then 10.5 ml 3 M glycine added and incubated for a further 5 min. Cells

were harvested by centrifugation at 6,000 x g for 2 min at 4 °C. The pellet was

washed with 25 ml of ice cold PBS and flash frozen by using liquid nitrogen. Using

61

the cryogenic grinder, cell samples were lysed by running at 30 Hz for 1 min 30 sec

for three cycles, cooling in liquid nitrogen between runs. The samples were

transferred to a 15 ml tube and 2.2 ml of IP buffer added containing 0.1 mg/ml

RiboShredder and a protease inhibitor tablet. Samples were divided into 400 μl

aliquots and Bioruptor for 30 sec on 30 sec off for 35 cycles. 12 µl samples were

separated by agarose gel electrophoresis to check that fragments ranged in size from

200-500 bp.

Once the size of the fragments was confirmed, samples were aliquoted into 2 ml

microfuge tubes and 25 μl of magnetic protein G beads (NEB) added, followed by

incubation for 90 min at 4 oC, rotating at 20 rpm. Beads were then immobilised

using a magnetic rack and the supernatant removed and discarded. The beads were

resuspended in 750 μl of chilled IP buffer and transferred to a new 1.5 ml tube and

incubated for 10 min at 4 oC rotating at 20 rpm. After incubating the beads were

immobilised again and the supernatant removed. The beads were resuspended in 1

ml of chilled IP salt buffer and incubated for 10 min at 4 oC rotating at 20 rpm. The

beads were immobilised again and the supernatant removed. The beads were

resuspended in 1 ml of chilled IP buffer with additional salt and incubated for 10

min at 4 oC rotating at 20 rpm. The beads were immobilised again and the

supernatant removed. The beads were resuspended in 1 ml chilled TE buffer pH 8

and incubated for 10 min at 4 oC rotating at 20 rpm. The beads were immobilised

again and the supernatant removed. The beads were resuspended again in 1 ml

chilled TE buffer pH 8 and incubated for 10 min at 4 oC rotating 20 rpm. The beads

were resuspended again in 100 μl Elution buffer and 5 μl of Riboshreader added,

then incubated for 30 min at room temperature. Samples were incubated at 65 oC

overnight to de-crosslink and the beads immobilised and the supernatant containing

the immunoprecipitated DNA was transferred to a new tube. This was incubated

with 5 μl of 10mg/ml proteinase K at 55 oC for 2 h. Samples were used for qPCR

and sent to The Genome Analysis Centre (Norwich) for sequencing, 50 bp single

end on the HiSeq 2500.

62

2.9. Electromobility shift assay (EMSA)

EMSA was used to analyse and detect DNA/protein complexes. DNA was

radiolabelled and mixed with protein before being loaded and run on a gel. The

detection of complex formation is known by comparing the bands of the

individually-run components with the combined components.

2.9.1. Generating radiolabelled DNA probe

Oligos were annealed to generate a probe and then purified from single stranded

DNA using the QIAgen gel extraction kit according to the manufacturer’s

instructions. Probes were radiolabelled, 100 ng probe, 1.11 MBq of γ[32P]-ATP, 1 μl

T4 polynucleotide kinase, 2 μl 10 x Kinase buffer, made up to 20 μl with nuclease

free H2O was incubated at 37 oC for 30 min. The probe was purified again by

QIAgen PCR purification kit according to the manufacturer’s instructions.

2.9.2. Running EMSA gel

Samples were prepared by adding 1 ng of γ-labelled DNA, 2 μl of 5x Binding buffer

and 2 μl of 6x loading dye and a specific amount of NAD+/H as well as the Rex

protein or 1μg herring-sperm DNA. This was incubated at room temperature for 20

min and then loaded onto a 6% TBE-polyacrylamide gel and run at 120 V for 1 h 20

min. The gel was then vacuum dried, exposed to a storage phosphor screen, then

analysed using a Typhoon phosphorimager (GE Healthcare).

63

3. Chapter III: Development of tools

Overview

Previously few genetic tools were available to assist in G. thermoglucosidans genetic

manipulation; an important aim was therefore to develop genetic tools for the

engineering of the organism. Until recently the introduction of DNA into G.

thermoglucosidans was achieved solely by electroporation for which the size of the

plasmid limited transformation frequency (Cripps et al., 2009). It is important that

vectors with the ability to tolerate large synthetic gene clusters are developed, as has

been the case for antibiotic biosynthesis cluster engineering in Streptomyces (Fong et

al., 2007). Electroporation is also costly due to the need for specialised equipment,

whereas conjugation does not (although it is slightly more time consuming). We

therefore set out to develop and optimise a conjugative system with improved

transformation frequency, especially for low copy, large or non-replicating

plasmids.

Conjugation is a method of direct horizontal gene transfer where an origin of

transfer (oriT) is nicked by a relaxase enzyme and one of the vector DNA strands is

transferred into the recipient cell (Lee & Grossman, 2007). Prior to this study, an

oriT fragment from the plasmid pSET152 was cloned into the MCS of the

bifunctional replicating plasmid pUCG18 (M. Jury & M. Paget, personal

communication) and conjugation between E. coli and G. thermoglucosidans TM444

was demonstrated but not properly quantified. The donor strain was E. coli S17-1, a

strain in which the conjugation apparatus of plasmid RP4 is integrated in the

genome. However, this strain methylates DNA and it was reported that G.

kaustophilus had a methyl-directed restriction system (Suzuki & Yoshida, 2012).

Simultaneously, work at TMO Renewables led to the construction of an alternative

conjugative plasmid pN109, in which the RK2 oriT fragment along with traJ from

pRK2013 (Kreuzer, Gärtner, Allmansberger, & Hillen, 1989) was inserted into

pUCG18 at a position that did not disrupt the MCS, which was therefore more

favourable for cloning. pN109 was subsequently used and adapted further to

generate a non-replicating plasmid for gene disruptions and fusions. Initial

conjugation experiments at TMO Renewables used tri-parental mating protocols to

64

bypass the methylation system. Tri-parental mating uses three bacterial strains

including a helper, a donor and a recipient strain. The helper strain caries the genes

required for DNA transfer and conjugation, the donor carries the plasmid for

transfer, and the recipient is the bacteria where the plasmid will be introduced into.

However this method, although successful, was overcomplicated and time

consuming. This chapter describes the development of a single donor strain along

with a simplified conjugation protocol that reduces the time involved from three

days to two. We also investigated the possible importance of a methyl-directed

restriction system in G. thermoglucosidans by developing a S17-1 ∆dcm donor strain

and testing its conjugation frequency against wt S17-1.

This chapter describes the development of strains, vectors and methodology for

efficient E. coli - Geobacillus conjugation. The expansion of the genetic tool kit by the

development of a conjugative method will aid academic and industrial research and

ultimately will be vital for the commercial success of G. thermoglucosidans as a

chassis for the production of biocommodities.

This work builds upon an early example of conjugation into a Geobacillus strain,

demonstrating the efficient conjugation from an E. coli donor strain to G.

kaustophilus (Suzuki & Yoshida, 2012). Since completing this research, efficient

conjugation between E. coli and G. thermoglucosidans was reported (Tominaga,

Ohshiro, & Suzuki, 2016). However, our research optimises the protocol,

making it far less time consuming. We also present a conjugative module for a

range of recently created shuttle vectors.

3.1. Preliminary evidence for E. coli to G. thermoglucosidans

intergeneric conjugation.

3.1.1. E. coli strains

The conjugative protocols first produced by TMO Renewables tried using BL21 or

JM109 as a donor strain, with HB101 (pRK2013) as a helper strain to conjugate

into G. thermoglucosidans recipient. JM109 lacks the E. coli K restriction system, and

BL21 has a non-functional dcm due to mutation; the lack of such methylation

systems were thought to enhance the success rate of conjugation as DNA was not

digested by the recipient. The plasmid pUCG16T in E. coli BL21 (dcm background)

was used as the conjugal donor and E. coli HB101 (pRK2013) as the conjugal helper

65

strain. In this system pRK2013 aided the transfer of pUCG16T from E. coli BL21

into G. thermoglucosidans. The purpose of this triparental system was to ensure that

conjugating DNA was not dcm methylated, thereby bypassing a suspected methyl-

directed restriction system. To reduce the complexity of the protocol and to

hopefully increase conjugation frequency it was an aim to develop and adapt E. coli

S17-1 in a biparental method (genotype: Tpr Smr recA thi pro hsdr hsdM+ RP4:2-

Tc::Mu::Kmr::Tn7 λpir). The conjugal transfer functions (RP4) of S17-1 are stably

integrated in the chromosome; this ensures that the helper plasmid (pRK2013) is

not co-transferred, which might otherwise complicate strain construction. However,

S17-1 methylates DNA and so initial studies were performed to see if, despite this,

the strain was suitable for conjugation into G. thermoglucosidans. Since the growth

stage of donor strains has an important effect on conjugation efficiencies, a key

benefit of using a single donor strain is that it removes the requirement to

coordinate growth of two donor strains and thereby provides more consistent

conjugation rates.

3.1.2. Initial biparental mating protocol

A biparental mating protocol was designed where G. thermoglucosidans and E. coli

were grow up as liquid cultures to an OD600 of 0.5, and then placed on a

nitrocellulose disk via vacuum pump in a ratio of 1:9. The nitrocellulose disk was

incubated on plates at 37oC over night to allow for conjugation to occur. The disk

was then washed to collect the cells, and the cells plated onto TSA + 12.5 µg/ml

Kanamycin and incubated at 60oC overnight to select for ex-conjugates. The

described protocol has undergone many alterations and optimisations throughout

this project. Our next aim was to streamline and optimise the bi-parental mating

protocol, refining the incubation periods, growth media, selection process, and

reducing the need for specialist equipment, thereby shortening the time

requirements and making the protocol more accessible. The optimised protocol can

be found in section 2.4.18.2. The main improvements were at stage 4 and 5 of

Figure 3.1 where the use of a nitrocellulose disc was discontinued and its function

of allowing transfer of cells from non-selective to selective media was replaced by

an antibiotic overlay. The conjugation period was shortened from overnight to ~2

66

hours and a period prior to antibiotic selection at 60 oC was found to lead to far

enhanced ex-conjugate recovery.

Figure 3.1 Flow diagram of original biparental mating conjugation protocol.

3.2. Growth media

Initially, to optimise conditions for E. coli LA growth medium was used during the

conjugation. However, no ex-conjugates were recovered, possibly because G.

thermoglucosidans was unable to survive the combined stress of sub-optimal medium

and temperature (data not shown). We therefore used TSA plates for further

experiments, which did not have any noticeable effect on the growth of E. coli and is

an optimal growth medium for G. thermoglucosidans. Divalent cations such as

magnesium have been shown to improve conjugation in some systems (Yu & Tao,

2010). Experiments were therefore performed to test whether magnesium improved

the conjugation. A concentration of 10mM MgCl2 more than doubled transfer

frequency and was therefore used in future conjugation experiments (Table 3.1).

1. Incoulate liquid media and incubate donor at 37 oC and recipiant cells at

60oC

2. Incubation continues until OD600 = 0.5

3. Mix donor and recipent cells, ratio of 1:9 and vacume pump onto

nitrocelulose disk

6. Single ex-conjegate colonys become visible

after over night incubation.

5. Harvest cells from nitrocellulose disk and plate onto selective (kan)

TSA media. Incubate overnight at 60 oC

4. Incubate nitrocellulos disk on TSA plate

overnight at 37 oC for conjugation

67

pN109 transfer frequency (recipient-1)*

Donor Strain methylase Gene

MgCl2 0 mM MgCl2 10 mM

S17-1

S17-1 ∆dcm

dam+, dcm+

dam+, dcm-

(7.1 ± 9.9) x 10-4

(1.2 ± 2.0) x 10-3

(1.9 ± 1.8) x 10-3

(3.0 ± 3.9) x 10-3

Table 3.1 Effect of magnesium supplementation on transfer frequency.*Transfer efficiencies are

expressed as the number of kanamycin-resistant transformants per total number of recipients. Results are expressed as the mean ± SD (MgCl2 0mM n=10, MgCl2 10mM n=5 )

3.3. Investigating the importance of DNA methylation on

conjugation frequency

Published work using Geobacillus kaustophilus (Suzuki & Yoshida, 2012) suggested

that DNA which is methylated by the dam dcm methylation system of a donor

strain such as E. coli, may be restricted. More specifically it appeared that dcm

methylated was particularly restricted, while dam-methylated DNA actually

transferred more efficiently (Suzuki & Yoshida, 2012). The donor strain S17-1 has

both dam and dcm methylases, therefore our initial aim was to construct an S17-1

dcm mutant; this was achieved using a λ red recombineering approach described in

Section 2.4.15 (Gust et al., 2004).

3.3.1. Construction of S17-1 ∆dcm::apr donor strain using a

REDIRECT approach

REDIRECT is a method of gene disruption using homologous recombination to

replace a target gene with an antibiotic resistance cassette (Gust et al., 2004). The

dcm gene in E. coli S17-1 was disrupted by replacing the dcm open reading frame

with an apramycin resistance gene (apr). A disruption cassette consisting of the

apramycin resistance gene (apr; aac(3)-IV) flanked by dcm-specific coding DNA was

amplified from the vector pSET152. Primers were designed to include the entire

68

aac(3)-IV gene including an upstream region of 130 bp, which was likely to include

the promoter. The 58nt primers included the DNA equivalent to the N-terminal

and C-terminal 13 amino acids (39 bp) of dcm at their 5’ ends and apr-specific DNA

at their 3’ ends. pIJ790 (λred inducible by arabinose, CmR) was introduced into S17-

1 and the λred system induced. The linear mutant allele was electroporated into

S17-1 (pIJ790) selecting for apramycin resistance. Several apramycin resistant

(AprR) colonies were isolated and screened for the mutant allele by colony PCR

(See section 2.2.3 col_S17-1_Dcm_F and col_S17-1Dcm_F ’). An E. coli S17-1 dcm

mutant strain was thereby successfully isolated. Colony PCR was performed to

distinguish mutants from non-mutants (see Figure 3.2 A); all isolates were found to

be mutants. To ensure that the dcm methylation system has been disrupted, a ScrF1

restriction enzyme digest profile of pN109 from each of the strains was compared

(see Figure 3.2 B). ScrF1 cleavage at restriction sites becomes blocked if DNA is

subject to either dam or dcm methylation. The loss of dcm methylation due to the

creation of a E. coli S17-1 dcm mutant strain is indicated, therefore, by an

increase/change in the restriction pattern created when plasmid pN109 is subject to

ScrF1 digestion.

69

Testing the effect of dcm methylation on conjugation frequency

A Mutant dcm amplified region = 1.2 kb. All colonys isolated were mutants. Wild type would

give a band size of 1.6 kb.

Figure 3.2 Agarose gel electrophoresis showing ScrFI restriction enzyme profile of

pN109 isolated from different E. coli strains. A Lane 1-4 isolated dcm mutant

colonies that underwent PCR, lane 5: DNA ladder. B Lane 1, DH5α; Lane 2, S17-1 ∆dcm::apr; Lane 3: S17-1 dcm+. Lane 4: DNA ladder. Note the appearance of smaller

fragments in the dcm mutant indicating that dcm methylation was not preventing ScrFI

digestion.

B

DNA bands of interest, indicating by an alternative digest profile difference in

methylation.

Co

lon

y

4

DH

5

S17-1

∆d

cm

Co

lon

y

3

Co

lon

y

2

S17-1

Co

lon

y 1

Lad

der

Lad

der

70

To test whether dcm methylation affects conjugation, experiments were performed

using E. coli S17-1 (pN109) or E. coli S17-1 Δdcm (pN109) and G. thermoglucosidans.

In addition, a non-replicating plasmid that integrates by homologous recombination

was tested: pSX700::∆rex is described in Section 4.3 and contains 1.3 kb of

homologous DNA either side of a ∆rex mutant allele. The transfer frequency of

plasmid pN109 from S17-1 Δdcm strain was ~1.7 fold higher than that from S17-1

(see Table 3.2), suggesting that dcm methylation might contribute to the degradation

of incoming plasmid DNA. However, it should be noted that if G. thermoglucosidans

possesses a methyl-directed restriction system, but that sites were not present on

pN109, we would not expect a difference in frequencies; i.e. other/larger

recombinant plasmids might benefit from the use of the Δdcm strain to a higher

degree. The non-replicating plasmid pSX700 showed a similar ~3 fold increase in

frequency (see Table 3.2), therefore the dcm mutant was routinely used in further

experiments.

donor Strain methylase gene

transfer frequency (recipient-1)*

pN109 pSX700::∆rex

S17-1 wild type S17-1 ∆dcm

dam+, dcm+

dam+, dcm-

(7.1 ± 9.9) x 10-4

(1.2 ± 2.0) x 10-3

(4.9 ± 2.9) x 10-6

(1.5 ± 1.5) x 10-5

Table 3.2 Transfer frequency of both replicating and non-replicating plasmids. *Transfer

efficiencies are expressed as the number of kanamycin-resistant transformants per total number of recipients. Results are expressed as the mean ± SD (pN109 n=10, pSX700 n=4 )

71

3.4. Construction of a non-replicating conjugative plasmid

It was anticipated that optimisation of the conjugation protocol would lead to

sufficiently high transfer rates to allow the use of a non-replicating plasmid for gene

disruption via homologous recombination. The non-replicating plasmid was

constructed by deleting a substantial fragment of DBNA from the repBST1 origin of

replication. repBST1 is an origin or replication derived from pBST22 and shares a

high similarity with the origin of replication from pBM300 (H. H. Liao & Kanikula,

1990). Alongside pN109, repBST1 is also utilised by the G. thermoglucosidans

PstI

EcoRV

pN109

MCS

Figure 3.3 The structure of pN109_oriT. The EcoRVand PstI restriction site locations on the

repBST1 origin of replication are indicated, used to render the plasmid non-replicating.

72

plasmid pUCG18 (Taylor, Esteban, & Leak, 2008). To disrupt the functioning of

repBST1 a 545 bp region of the origin of replication was deleted from pN109 by

digestion with PstI and EcoRV (see Figure 3.3). The vector was then re-circularised

by blunt end ligation and the resulting vector named pSX700 (Figure 3.4).Vector

pSX700 was tested to ensure it was unable to replicate in G. thermoglucosidans. The

non-replicating conjugative pSX700 plasmid was then successfully used to promote

integration of a cassette into the genome of Geobacillus thermoglucosidans via

homologous recombination.

pSX700

MCS

Figure 3.4 Non-replicating conjugative plasmid pSX700.

73

3.5. Development of and optimising a rapid conjugation protocol

3.5.1. Conjugation temperature and time optimisation

The donor and recipient organisms have very different optimal growth

temperatures. The G. thermoglucosidans optimal growth temperature is 60oC with a

tolerance down to 37 oC, whereas the E. coli optimal growth temperature is 37oC

with a tolerance up to 45 oC. It was therefore no surprise that the duration and

temperature of conjugations were key variables that had a dramatic effect on

transformation frequency.

Initial experiments indicated that conjugations should be performed at 37oC to

ensure E. coli can express all necessary transfer functions; no ex-conjugates were

recovered when a conjugation temperature of 45 oC was trialled (data not shown).

Although at first this incompatibility in temperature was considered a disadvantage

it does prove a useful additional selection to eradicate E. coli growth later in the

protocol, and prevents G. thermoglucosidans from growing during the early stages of

conjugation, thereby generating siblings.

The length of time for which conjugation was allowed to proceed was also found to

be significant (see Table 3.3); optimum frequencies were obtained when

conjugation was performed overnight (17 h) at 37oC (Table 3.3). However,

shortening this period to 2 h only reduced frequency by a factor of ten. A further

Conjugation duration (hours) pN109 Transfer frequency (recipient-1)*

1

(2.4 ± 1.1) x 10-6

2 (1.2 ± 2.6) x 10-5

17 (3.8 ± 3.2) x 10-4

Table 3.3 Conjugation duration optimisation. *Transfer efficiencies are expressed as the number of kanamycin-resistant transformants per total number of recipients. Results are expressed as the mean ± SD, n=3. All recovery times before kanamycin selection

were a duration of 1 h at 60oC.

74

reduction of conjugation incubation time to 1 h diminished frequency again by a

factor of ten.

The 1 h conjugation period achieves sufficient frequency of 2.4 x 10-6 if a replicating

plasmid is used, and shortens the protocol time by the largest amount, however a

conjugation period 2 h is almost 10 times more effective and is still achievable in

one day and so was the preferred choice. Non-replicating plasmids that integrate via

homologous recombination require a high conjugation frequency and so 3 h is

routinely used (see Table 3.4), and in the case of pSX700::Δrex was found to have

an ample frequency of 2.2 x 10-6. When conducting the protocol, it can be expected

that ~300 colonies will be found to grow on a single plate after transforming the

replicating plasmid, and ~30 for the non-replicating plasmid.

Conjugation duration (hours) pSX700::Δrex Transfer frequency (recipient-1)*

3 (2.2 ± 8.5) x 10-6

17 (1.4 ± 1.9) x 10-5

Table 3.4 Conjugation duration optimisation of non-replicating plasmid. *Transfer efficiencies

are expressed as the number of Kanamycin-resistant transformants per total number of recipients. Results are expressed as the mean ± SD n=3. All recovery times before kanamycin selection were a duration of 1.5 hours at 60oC.

3.5.2. Recovery Period

An important step in any transformation is the recovery period during which the

recipient strain is given time to express the antibiotic resistance gene used for

selection. The plasmids used in this work contained a kanamycin resistance gene,

and selection for ex-conjugates took place at a concentration of 12.5 µg/ml

kanamycin. In early protocols before optimisation, cells were plated out and

immediately incubated after a period of conjugation on TSA + 12.5 µg/ml

kanamycin plates. However, this resulted in a near zero conjugation success rate. A

period of recovery was introduced into the protocol to allow the ex-conjugate G.

thermoglucosidans cells to recover from temperature stress.

In typical bacterial transformation experiments, 1 h at the recipients’ optimal

growth temperature is usually sufficient before selection is applied. However, in this

75

case the recipient G. thermoglucosidans must not only begin to express the resistance

gene, it also must first recover from a stressful incubation period at 37oC. For this

reason, we explored and optimised the period of recovery needed.

If the 60 oC recovery duration was increased from 30 mins to 90 mins a 30-40 fold

increase in frequency was achieved (see Figure 3.5). To ensure this increase was not

due to the overlay redistributing G. thermoglucosidans cells after they had undergone

division, a control was set up. For the control, mixed donor and recipient cells were

plated onto a non-selective medium (note 10-3, 10-4 and 10-5 cell dilutions were used)

and grown at 37 oC for 2 h, then allowed to recover for 30 min at 60 oC. Half the

plates were then overlaid with 500 µl of liquid media only (representing a

kanamycin overlay), and all plates then incubated at 60 oC overnight. The

difference in cell count was measured between those plates which underwent the

overlay and those that did not. It was found that the overlay did not have the effect

of redistributing cells so as to superficially increase colony numbers. The opposite

effect was in fact seen, whereby the overlay disturbed cell growth and reduced cell

numbers slightly. We therefore believe the overlay is not contributing to

exaggerated ex-conjugate number, but instead reduces overall frequency of the ex-

conjugates recovery.

76

The optimal time for the transfer of pN109 from the donor to the recipient was

found to be 2 h, and the recovery time 90 min, giving a transformation frequency of

2.4 x 10-5. These time parameters can be adjusted depending upon the plasmid for

transfer and the number of ex-conjugates needed. For example, when using a non-

replicating plasmid, or for generating a mutant library, a higher frequency will be

required.

3.5.3. Description and Comparison of Protocols

The biparental mating protocol was optimised to reduce the time involved and the

specialist equipment needed. The recovery of ex-conjugates was optimised and a

S17-1 ∆dcm strain constructed to circumvent methylation restriction. A non-

replicating plasmid (pSX700) was also produced to integrate by homologous

Figure 3.5 Transformation efficiency rates of pN109 with variation in time. Comparison of

transformation frequency rates of pN109 S17-1 Δdcm for varying conjugation durations (1 and 2

hours) and 19955 ex-conjugate 60 oC recovery durations (30, 60, 90 minuets).

77

recombination, and was use to create a Δrex strain. Figure 3.6 is of an overview of

the different protocol stages. It differs from the original conjugation protocol at

stages 4, 5 and 6. The introduction of an overlay step is the main addition, which

replaced the need for a nitrocellulose disk which was used to transfer cells from

non-selective, to selective media.

Figure 3.6 Overview of optimised conjugation protocol.

3.6. Adapting SEVA modular shuttle vectors

Recently, a range of G. thermoglucosidans - E.coli shuttle vectors were created as part

of a modular toolkit by the Ellis group (Reeve et al., 2016). The plasmids were

based on the ‘Standard European Vector Architecture’ (SEVA) model of

interchangeable parts, allowing deconstruction and reconstruction of plasmids,

giving the flexibility to exchange functional parts according to the needs of the user.

These vectors contain a variety of reporter genes, including a gfp, which is applied

in Chapter 6 to assess gene expression. However, the plasmids lack oriT and so

require electroporation to transform Geobacillus. It was therefore decided to

incorporate oriT into these plasmids thereby permitting high frequency conjugation.

1. Incoulate liquid media and incubate donor at 37 oC and

recipiant cells at 60oC

2. Incubation continues until OD600 = 0.5

3. Mix donor and recipent cells, ratio of 1:9 and plate

onto TSA.

6. Overlay with antibiotic selection (kan) to select for ex-

conjugets and incubate overnight at 60 oC.

5. Move TSA plate to 60 oC incubator for 90 min for stress

recovary.

4. Incubate TSA plate for 2 h at 37 oC for conjugation.

78

Our aim was to make a modular oriT component, designed to be compatible with

the existing G.thermoglucosidans / E.coli shuttle vectors and the current range of

parts. The conjugative module was designed to conform, where possible, to the

SEVA format (see Figure 3.7). This requires the absence of particular internal

restriction enzyme sites (HindIII, PstI, XbaI, BamHI, SmaI, KpnI, SacI, SalI, EcoRI,

SfiI, SphI, AatII, AvrII, PshAI, SwaI, AscI, FseI, PacI, SpeI, SanDI and NotI) along with

flanking sites that meet the SEVA format and ensure modular compatibility.

The conjugative module was based on the broad host rage RK2 family oriT and its

sequence synthesised by Life Technologies as a double stranded GeneArt string of

792 bp in length. EagI and NotI sites were placed at the ends of the fragment to

allow it to be subclonned at the NotI site of the shuttle vectors, specifically pG1AK-

sfGFP (as this vector was to be used for further experiments due to it containing

gfp). The oriT double stranded DNA fragment was first cloned into the EcoRV site

Figure 3.7 The elementary design of a SEVA plasmid. Each functional part is divided by rare restriction sites AscI, PacI, SpeI, SwaI, PshAI and FseI. Terminators T1 and T0 flank the cargo region and aid in plasmid stability. An oriT permits the conjugative transfer of the plasmid between the donor strain E.coli and a recipient (Silva-Rocha et al., 2013).

79

of pBlueScript II SK+, then subcloned as a EagI-NotI fragment into the NotI site of

pG1AK-sfGFP (which had been digested with EagI to give compatible overhangs,

see Figure 3.8). When ligated, the NotI site would be recreated at only at one end,

and a EagI would form at the other so to conform to the SEVA format (see Figure

3.9).

80

OriT

NotI EagI

GC GGCCGC--//--TC GGCCG

CGCCGG CG--//--AGCCGG C

p11AK2

Figure 3.8 Ligation of OriT into p11AK2. The sticky end overhangs after EagI digestion to

exercise the oriT fragment, once ligated to the NotI cut vector backbone, resulted in the recreation of NotI and EagI sites. This was designed so to conform to SEVA.

Figure 3.9 Vector map of p11KA2_oriT_Rex. An oriT has been ligated into the p11AK2 NotI

site (which had been cut with EagI).

81

3.6.1. Testing conjugation frequency of pG1AK-sfGFP_oriT

The conjugation frequency of pG1AK-sfGFP_oriT was compared with that of

pN109 using a range of incubation times (see Table 3.5). This revealed that the

pG1AK-sfGFP_oriT transfer frequency was ~10-fold lower than that of pN109.

The reason for this is unclear but is unlikely to be due to differences in size (pN109

is only 298 bp a smaller plasmid than pG1AK-sfGFP_oriT). Other possible

explanations are differences in copy number and in host restriction.

Transfer frequency (recipient-1)*

Plasmid

60 mins (recovery duration) 90 mins (recovery duration)

pN109

(2.4 ± 1.1) x 10-6

(6.5 ± 1.5) x 10-6

pG1AK-sfGFP_oriT

(2.9 ± 0.9) x 10-7

(5.1 ± 1.5) x 10-7

Table 3.5 Transfer efficiencies with varying recovery times. Transfer efficiencies are expressed as the number of Kanamycin-resistant transformants per total number of recipients. Results are expressed as the mean ± SD n=3. All conjugation times were a duration of 1 hour at 60oC. pG1AK-sfGFP_oriT = 6435 bp, pN109 = 6137 bp. Donor

strain used was S17-1 Δdcm.

3.7. Attempts to develop an integrative conjugative system

The use of chromosomal integration of biosynthetic modules avoids issues

associated with multi-copy vectors such as plasmid segregation and structural

instability. It also removes the need for maintaining antibiotic selection, which is

particularly important for industrial applications. Currently, only homologous

recombination has been used to integrate new genes into the G. thermoglucosidans

genome. This is a laborious two-step process not suited to the screening of multiple

constructs. Flanking regions are generated which are homologous to the site of

integration. Recombination of this type is an event reliant on probability, its

82

frequency of occurrence proportional to the length of flanking regions and the

recombinogenic tendency of the specific genome region (Marshall Stark et al.,

1992). Therefore a site-specific DNA integration system to rapidly engineer G.

thermoglucosidans is of particular interest and was partly developed.

Site-specific recombination relies upon DNA-binding recombinases, and is far more

efficient than homologous recombination (Lyznik et al., 2003). It evolved as a

mechanism by which mobile elements including viral DNA could integrate into

host genomes (Groth and Calos, 2004). The bacteriophage φC31 is a natural virus

of Streptomyces coelicolor; upon infection the circular DNA integrates into a specific

small site in the genome via site-specific integration. The natural φC31 site-specific

recombination system has been adapted for the generation of several integrative

vectors in Streptomyces (Bierman et al., 1992). Furthermore, it has been adapted for

use in higher organisms including fly and mammalian cells (Bateman et al., 2006,

Marshall Stark et al., 1992, Sadowski, 1986). The system consists of three main

components: integrase enzyme, attachment phage (attP) and attachment bacterial

(attB) sites. The integrase enzyme cuts DNA at specific att sites and initiates

recombination: one attachment site on the host genome, and the other on the

plasmid. These are brought into close proximity by protein-protein-interactions

mediated by the integrase enzyme, cleavage occurs of the plasmid backbone and

genome after which strands undergo a strand exchange reaction and ligate together

(Coates et al., 2005, Fogg et al., 2014).

In this work, it was aimed to generate a conjugative-integrative system based on

φC31 for use with G. thermoglucosidans. Bioinformatic searches showed no natural

attB attachment site present in the G. thermoglucosidans genome, which therefore

necessitated the artificial engineering of a site. An integrative vector containing the

φC31integrase (int) and attB sites was constructed, the vector backbone based upon

the non-replicating conjugative plasmid pSX700, allowing manipulation in E. coli,

and transfer by conjugation. It was aimed to integrate the attP site in the host

genome rather than attB, thereby preventing the (unlikely) infection of the strain

with a related bacteriophage.

83

3.7.1. Integrative Vector

The integrative plasmid includes φC31 system parts cloned into pSX700 for E. coli

to G. thermoglucosidans conjugation. The necessary parts were cloned into pSX700 in

two stages. Firstly the minimal attB integration site (70bp) was determined using

the literature (Bateman & Wu, 2008) and synthesised as a single fragment including

the chosen promoter for the integrase gene (pup3 – a constitutive promoter based

on pUP constitutively expressed uracil phosphoribosyl transferase gene of G.

thermoglucosidans and has been found to be functional in both E. coli and G.

thermoglucosidans (Reeve et al 2016)) and homologus flanking regions for Gibson

assembly into pSX700 (see Figure 3.10). XhoI and SpeI restriction sites were placed

in between the promoter and the attB site so that the integrase gene could be ligated

in at a later stage. The fragment was produced by Life Technologies, as a double

stranded GeneArt string and was directly Gibson assembled into NotI cut pSX700

to create pSX700_attB. The second stage was to PCR amplify the φC31 integrase

gene (int) from pSET152 using primers (see section 2.2.3 Int_F and Int_R) with

restriction sites XhoI on the forward, and SpeI on the reverse for later ligation. The

reverse primer was designed to also contain a terminator sequence. Although the

φC31 integrase gene is unlikely to be thermostable at 60oC this is unimportant since

conjugation with E. coli is carried out at 37oC. After amplification the fragment was

purified and cloned into EcoRV site of pBlueScript, sequenced to check for errors

(one silent mutation was found), and cut out as an XhoI and SpeI fragment. It was

then ligated into pSX700_attB to create pSX700_attB_int (Figure 3.11).

Conjugation experiments set up with S17-1 dcm (pSX700_attB_int) and G.

thermoglucosidans 11955 resulted in no ex-conjugants, as expected since a genomic

attP site is not present in the genome.

84

Figure 3.10 Vector map of the pSX700 and linier synthesised attB fragment. The NotI site

where the pSX700 vector was linearized for a synthesised fragment to be Gibson assembled in, is indicated. The synthesised fragment contains: 40 bp pSX700 homologues regions at either end; a pup3 constitutive promoter for integrase expression; XhoI and SpeI sites for ligation of the PCR amplified integrase gene; a 70 bp attB site.

pSX700_OriT

MCS

XhoI SpeI

homologous Pup3 homologous attB

NotI

homologous

85

XhoI SpeI

int

A

pSX700_OriT_attB XhoI

SpeI

MCS

B

pSX700_OriT_attB_int

MCS

Figure 3.11 Integrative vector construction. A) Vector map of pSX700 and the int fragment (dull yellow, the terminator a darker yellow) to be ligated into the XhoI and SpeI restriction sites. B) Final map of the integrative vector pSX700_attB_int.

86

3.7.2. Genomic attachment site

During this study several alternative genomic locations for the attP attachment site

were chosen, and attempts made in each case to generate a stable recombinant

strain; however, none were successful. The genomic sites selected were a mutant ldh

gene, and two genes that provide the possibility of efficient counter selection: pyrE

and upp.

The orotate phosphoribosyltransferase (pyrE) gene as an

attachment site

In order to rapidly generate strains that have an integrated attP site, it would be

beneficial to place the attachment site in a gene, the disruption of which could be

selected for (counter selection). The orotate phosphoribosyltransferase (pyrE) gene

has been used as counter selectable marker in many organisms including in yeast,

Bacillus and Clostridium (Boeke, La Croute, & Fink, 1984; Haas, Cregg, & Gleeson,

1990; Suzuki, Murakami, & Yoshida, 2012; Tripathi et al., 2010). The pyrE gene is

part of an operon integral for the pyrimidine biosynthetic pathway, and it catalyses

one of the last steps in this process and disruption of this gene prevents the

conversion of orotic acid to UMP and produces uracil auxotrophs (Dong & Zhang,

2014). Disruption of pyrE expression also prevents the conversion of 5-fluoroorotic

acid (5-FOA) into toxic 5-fluoro-UMP. Therefore the replacement of pyrE with attP

by homologous recombination should result in 5-FOA resistance. However, as pyrE

is essential for growth on pyrimidine-deficient medium, uracil is incorporated into

5-FOA selection media.

87

pyrE gene disruption

The pyrE is located within an operon with orotidine 5'-phosphate decarboxylase

(pyrF) and dihydroorotate dehydrogenase (pyrD) (see Figure 3.12). A 2.02 kb

fragment of DNA including the centrally located pyrE gene was amplified by PCR

using G. thermoglucosidans chromosomal DNA as template and primers that

incorporated flanking BamHI sites. Following cloning into pBlueScript, inverse

PCR was performed to delete 444 bp of pyrE coding region, replacing it with the 60

bp attP site. The pyrE_att construct was subcloned into pSX700 and transformed

into S17-1 Δdcm. pSX700_ pyrE_att (see figure 3.13) was then conjugated into G.

thermoglucosidans and ex-conjugants were selected using kanamycin, thereby forcing

genomic integration by homologous recombination. Colonies were picked, purified

as kanamycin resistant clones, grown in TSB to allow time for the second crossover

event to occur, then plated to minimal media containing 50 mg/l 5-FOA,

supplemented with 10 mg/l uracil. This was predicted to select for the second

crossover event forcing the loss of the wild-type pyrE gene and the generation of a

stable mutant allele that includes a central attP site. Cell dilutions were also plated

onto TSA only, and TSA with kanamycin; this was done to determine the

phenotypic makeup of the liquid culture. The results indicated (see Table 3.6) that

few second crossovers arose during the non-selective liquid incubation period. The

7 5-FOA resistant colonies were streaked onto TSA with kanamycin selection, MM

with 5-FOA and uracil, and TSA with 5-FOA. Of the seven potential pyre mutant

colonies only two were found to be both kanamycin sensitive and 5-FOA resistant.

As a control measure WT G. thermoglucosidans was also streaked alongside the

potential pyrE mutants.

pyrD pyrF pyrE

Figure 3.12 The genes surrounding pyre. The pryE gene encodes an orotate phosphoribosyltransferase, pyrF encodes orotidine 5'-phosphate decarboxylase and pyrD encodes

a dihydroorotate dehydrogenase. Genes overlap one another and exist in the same operon.

88

Dilutions of cells grown in TSB 60oC

with no selection & no Uracil addition.

TSA only TSA +

Kan

MM + 5-

FOA +

Uracil

10-2 dilution 10,000 + 10,000+ 7

10-4 dilution 1384 1104 0

10-5 dilution 352 83 0

Table 3.6 Serial dilution of TSB cell coulters grown from first crossovers for phenotypic analysis.

pSX700_pyrE_attP

BamHI

BamHI

Figure 3.13 Vector map pSX700_pyrE_attP. pyrE flanks were ligated into the BamHI site of pSX700. The attP site was constructed by inverse PCR and is situated in-between the pyrE flanks

so to be inserted into the genome of G. thermoglucosidans via homologous recombination.

89

G. thermoglucosidasius

strain type TSA MM +

FOA

MM + U TSA +

Kan

MM +

FOA +

U

ΔpyrE::attP ++ - + - +

11955 ++ - + - ++

11955 first crossover ++ - + + ++

Table 3.7 Comparison of G. thermoglucosidasius strain phenotypes. Vigorous growth is indicated

as ++, moderate +, and no growth as - .

Unexpectedly the WT strain was able to grow on 5-FOA selective media (see Table

3.7). The results indicated the phenotypes of both the potential pyrE mutants and

WT strains was indistinguishable. The potential pyrE mutants were tested via

colony PCR, the results indicated that no mutants were isolated (results not

shown). Several attempts were made to isolate a potential pyrE mutant using

variations of the method previously outlined. A range of 5-FOA concentration were

tested (1-200 µg/ml) and uracil supplementation (1-10 µg/ml), despite this we were

unable to differentiate between WT and potential pyrE mutants using phenotypic

analysis. Literature suggests this may be due to uracil supplementation and its

inhibitory effect on the expression of the operon containing pyrE. A regulator of the

pyrimidine biosynthetic genes, pyrR was found, able to sense UMP and UTP and

inhibit gene expression in Bacillus (Dong & Zhang, 2014). UMP is produced from

uracil by the protein encoded in upp gene. PyrR works as a negative regulator and is

a mRNA-binding attenuator (Turner, Bonner, Grabner, & Switzer, 1998). A

bioinformatics search revealed a pyrR homologue in G. thermoglucosidans with a 73%

identity.

Our results indicated that counter selection was not possible as the necessary uracil

enriched media caused the pyr operon to be inhibited and WT strains were therefore

no longer susceptible to 5-FOA selection. Compounding this issue was that any

pyrE mutant strain was likely slower growing than the WT. For these reasons the

90

counter selection system failed and in fact pyrE mutants/ genomic attP site

integration was more difficult to isolate than they would have been otherwise by

mutant screening and colony PCR.

The uracil phosphoribosyl-transferase gene (upp) as an

attachment site

Based on successful work in Bacillus subtilis (Fabret, Ehrlich, & Noirot, 2002) it was

decided that the attP site would be placed into the uracil phosphoribosyl-transferase

gene (upp) of G. thermoglucosidans. The uracil-phosphoribosyl-transferase (UPRTase)

enzyme converts uracil into UMP, allowing the cell to use exogenous uracil

(Neuhard, 1983; Nygaard, 1993). It also converts the pyrimidine analogue 5-

fluorouracil (5-FU) to a toxic product and can therefore be used as a counter

selectable marker. The upp gene is the only gene present within the operon.

upp gene disruption

Two 980 bp flanks were designed based on the upp gene and PCR amplified from

G. thermoglucosidans using primers of 38 nucleotides in length, designed with a 20

nucleotide overlap with the next fragment in the Gibson assembly (See Section

2.2.3 UPP-LFlank_Fwd, UPP-LFlank_Rev , UPP-RFlank_Fwd , UPP-

RFlank_Rev). The attP site was PCR amplified as a 300bp fragment, using

pSET152 as a template using primers designed for downstream Gibson assembly

(See Section 2.2.3, attP_pSET1_Fwd, attP_pSET1_Rev). In total 4 fragments were

assembled in a single reaction including the pSX700 (BamHI cut) backbone, both

the left and right upp flanks and the attP site amplified from pSET152 (see Figure

3.14).

91

Upp flank left Upp flank right

BamHI

att P

pSX700 (BamHI cut)

Figure 3.14 representation of the overlap of fragments to be assembled via a signal Gibson reaction. Overlaps were a minimum of 20 nt in length.

pSX700_pyrE_attP

BamHI

Figure 3.15 Vector map pSX700_upp_attP. The attP site, along with upp flanks were Gibson assembled into the BamHI site of pSX700.

92

A upp::attP allele was generated by using Gibson assembly cloning the fragment

into the BamHI site of pSX700 (Figure 3.15). The assembly was checked via

diagnostic digest before transforming into S17-1 ∆dcm strain and conjugating into

G. thermoglucosidans. First crossover recombinants were selected for by kanamycin

resistance and then subcultured in non-selective TSB media, giving time for the

second crossover event to occur. The culture was plated to TSA containing 100

µg/ml 5-FU to select for the second crossover. The time taken for colonies to grow

to a visible size was 48 h compared to the usual ~6 h for a WT strain. Three KanS 5-

FUR colonies were isolated after screening for kanamyacin sensitivity. The KanS 5-

FUR colonies were identified as being possible mutants, and one was confirmed by

colony PCR. Despite attempts made to conjugate into the G. thermoglucosidans

Δupp/attP strain using the vector pSX700_attB to test if the integrative system was

functional, no ex-conjugates were isolated. This may be due to the weakness of the

G. thermoglucosidans Δupp/attP strain which takes 48 h to grow to a visible size and

may not cope with the temperature (37 oC) at which conjugation occurs. The

negative result could also be due to the integrative system being non-functional.

3.8. Discussion and future work

The genetic engineering tools available for use with G. thermoglucosidans have,

during this project, been expanded. The development of a conjugative system for

use in G. thermoglucosidans represents a major addition to the genetic tool kit,

enabling reliable and high frequency transformation of replicating and non-

replicating vectors. The optimisation of the rapid conjugation protocol has also

made this form of transformation advantageous when compared to electroporation

in terms of ease and time duration. It also eliminates the need for specialist

equipment. A number of vectors are now available for use with this protocol,

including a non-replicating vector and an adapted modular shuttle vector

component, compatible with a range of plasmids adopted by several research

groups (Reeve et al., 2016).

The production and use of a S17-1 ∆dcm::apr donor strain did improve conjugation

frequency. Previously published work using G. kaustophilus demonstrated a 10 fold

improvement in transformation efficiency when using an E. coli Δdcm donor strain

93

(Suzuki & Yoshida, 2012). In comparison a lesser improvement of between 1.7-3

fold was recorded for G. thermoglucosidans using a S17-1 ∆dcm::apr donor strain

(depending on the DNA conjugated). Whilst work here demonstrates the

importance of DNA methylation on transformation frequency, investigations into

the exact nature of this could be carried out. Perhaps an alternative methylation

system to the dcm also plays a role and investigation of this could help to enhance

transformation frequency further.

Temperature during the conjugation protocol was found to play a key role in a

successful outcome. E. coli and G. thermoglucosidans optimal growth temperatures

differ substantially. By regulating growth temperature carefully, switching between

37 oC and 60 oC, optimising each step of the protocol, what at first seemed a major

hindrance became an advantage. A balance was struck between allowing for a long

enough incubation period at 37 oC for conjugation to occur, but not too long as to

negatively affect G. thermoglucosidans. A recovery period for G. thermoglucosidans at

60 oC was introduced, before the antibiotic selection was applied. By performing the

conjugation at 37 oC, optimal for E. coli, the overgrowth of G. thermoglucosidans was

halted, allowing for an accurate measurement of transformation frequency. The

switch to G. thermoglucosidans’s optimal temperature of 60 oC after conjugation

killed the E. coli cells and acted as a selection method. It was found that pre-

warming of all equipment and media was key to successful conjugation, as G.

thermoglucosidans was especially sensitive to cold shock, even at room temperature.

One of the many optimisations made to the conjugation protocol was the addition

of an antibiotic overlay, which replaced the need for nitrocellulose disks.

Nitrocellulose disks were used in early protocols to aid the transfer of cells from

TSA plates to TAS + kanamycin selection plates. Disks were removed from the

TSA plate and cells washed off, then resuspended and plated onto selective media.

During this wash step the cells were disturbed which led to a reduction in ex-

conjugate recovery. Furthermore, this method required specialised equipment to

place the cells on the nitrocellulose disk. It also exposed G. thermoglucosidans to

room temperature (known to be harmful, resulting in cell death) whilst the culture

was vacuum pulled through the discs, trapping the cells on the nitrocellulose

matrix. To eliminate the need for nitrocellulose disks, an overlay step was added to

the protocol. After conjugation, instead of transferring cells by the method outlined

94

above, the TSA plates containing ex-conjugates (without a nitrocellulose disk) were

overlaid with a kanamycin-TSB dilution. This gave good results, was quicker to

perform and was less expensive.

The work we have carried out to develop an integrative conjugative system will

hopefully be of use to guide future research. It is as yet unclear if the integrative

system is viable in G. thermoglucosidans. The integrative vector remains to be fully

tested for its capability to express the integrase machinery at the correct time point.

The greatest task remains the engineering of an attachment site into the genome of

G. thermoglucosidans. This could be achieved by simple homologous recombination

and screening methods for the loss of antibiotic resistance.

It is likely that the creation of a G. thermoglucosidans ΔpyrE strain and its use as a

counter selectable marker may still be possible despite the limitations of this project.

If the pyrR gene was to be inactivated by deletion it would no longer be able to act

as a negative regulator of pyrE expression (by sensing UMP and UTP levels which

were elevated by the enrichment of media with uracil) and it might become possible

to differentiate the phenotype of WT and ΔpyrE G. thermoglucosidans strains.

95

4. Chapter IV: Analysing the Rex regulon

Overview

Rex is a transcriptional repressor that senses cellular NAD+/NADH levels and

regulates specific genes, often key to fermentation and anaerobic respiration, so to

maintain redox balance. During periods of environmental oxygen abundance,

aerobic respiration functions as the main process to produce energy, and NADH is

constantly recycled to NAD+, maintaining redox balance. During oxygen limited

growth conditions however, NADH builds up, as the terminal electron acceptor,

oxygen, is scarce. Just as the lack of oxygen limits the electron transport chain and

ATP synthesis, the lack of NAD+ becomes a rate limiting step for the glycolysis

pathway. To cope with these conditions bacteria often switch to nitrate as an

alternative terminal electron acceptor. Soil-dwelling bacteria such as Geobacillus

thermoglucosidans often encounter oxygen limited environments and are able to

respire anaerobically, or switch to fermentation if no terminal electron acceptor is

available (Bueno, Mesa, Bedmar, Richardson, & Delgado, 2012). The Rex regulon

has previously been described in a wide variety of Gram positive bacterium and

found to be a regulator of genes involved in fermentation, glycolysis, NAD

biosynthesis, and general energy metabolism, as well as some miscellaneous and

uncharacterised genes (D. A. Ravcheev et al., 2012).

Rex is of particular interest during this project due to its key role in regulating

fermentative genes. Understanding and utilising Rex to optimise the industrial

potential of G. thermoglucosidans is an important aim. To gain insight into the Rex

regulon a 3xFLAG tagged Rex strain was constructed for use with ChIP-

sequencing, a technique able to map all Rex binding sites within the G.

thermoglucosidans genome. In addition to this, a rex mutant strain was constructed

and analysed for effects of deregulation. The Δrex strain was predicted to experience

major loss in the control of fermentation genes. Results from the ChIP-seq were

used to validate potential members of the Rex regulon using qRT-PCR

experiments. This information has allowed us to gain a sophisticated knowledge of

the Rex regulon. qRT-PCR data suggested that other regulators might also act on

Rex targets, producing unexpected variations in transcript abundance, in response

96

to changes in oxygen availability. Fermentative genes such as adhE (alcohol

dehydrogenase) were shown to be deregulated in a Δrex strain, although there was

surprisingly small changes in overall fermentation products, as shown by HPLC.

4.1. The genetic context of rex

G. thermoglucosidans rex is a 639 bp gene located within a predicted operon

consisting of four genes (see Figure 4.1). rex is the second gene of the operon.

Upstream of rex is the first gene of the operon (ATO13_04535; gene numbering for

G. thermoglucosidans DSM2542 is used throughout this thesis) which encodes a

molybdenum cofactor biosynthesis protein C. Molybdenum cofactor is utilised by

many enzymes to catalyze redox reactions and is found universally in bacteria,

archaea and the eukaryotic domains of life (Bittner & Mendel, 2010). The genes

downstream of rex encode preprotein translocase subunits (AOT13_04545 and

AOT13_04550), part of a sec-independent selective twin-arginine translocation

(Tat) system (Berks, Sargent, & Palmer, 2000). The Tat system has been identified

as commonly transporting redox enzymes (Weiner et al., 1998). The rex gene in B.

subtilis is also arranged within a similar genomic context, with TatAy and TatCy

positioned downstream of rex. The TatAy – TatCy pathway is dissimilar from other

systems, such as the TatAd–TatCd pathway in E. coli (Goosens, Monteferrante, &

van Dijl, 2014). The TatAd–TatCd pathway was capable of translocating the

TatAy–TatCy-dependent substrate EfeB, however the TatAy – TatCy pathway was

not capable of translocating the TatAd–TatCd dependent PhoD (Eijlander,

Jongbloed, & Kuipers, 2009). Differences were also found in expression profiles:

TatAd and TatCd are expressed only in low phosphate conditions whereas TatAy

and TatCy were constitutively expressed in low phosphate conditions (Goosens et

al., 2014).

97

AOT13_04535 AOT13_04540 AOT13_04545 AOT13_04550

4.2. Identification of Rex targets by ChIP-seq analysis

To fully understand the biological role of Rex we used chromatin

immunoprecipitation assay in combination with a DNA sequencing (ChIP-

Sequencing), an approach that could globally map all Rex target binding sites

within the genome of G. thermoglucosidans. This revealed the potential regulon and

also helped identify Rex consensus binding motif using ChIP-Seq enriched DNA

regions.

The ChIP-Sequencing technique involves chemically crosslinking DNA binding

proteins to chromosomal DNA in vivo, and then selectively immunoprecipitating

the protein of interest — along with the bound DNA — using a specific antibody.

The co-immunoprecipitated DNA is then sequenced and mapped to the genome

revealing binding sites and a qualitative view of relative occupancy. An epitope tag

can be used to modify the DNA binding protein if no poly- and mono-clonal

antibodies against the protein are available. In this instance the addition of a

3xFLAG tag to the C-terminus of Rex was used to allow immunoprecipitation with

an anti-FLAG antibody.

4.2.1. Construction of a 3xFLAG tagged rex strain

The addition of a triple FLAG tag to the terminus of Rex was achieved in a number

of stages that included cloning steps for assembly and then subsequent integration

moaC rex TatA TatC

Figure 4.1 Organisation of the predicted G. thermoglucosidans DSM2542 rex (AOT13_04540)

operon. rex is the second gene in the four-gene operon. The start codon of rex overlaps with the last three codons of the first gene, molybdenum cofactor biosynthesis protein C (moaC)

suggesting translational coupling. The gene downstream, preprotein translocase subunit TatA (tatA) is separated by 15 bp. Preprotein translocase subunit TatC (tatC) is 196 bp downstream of

tatA.

98

into the genome to replace the WT rex gene. Firstly, the rex gene and surrounding

region was amplified from chromosomal DNA using (Grex_reg_F and

Grex_reg_R) primers containing BamHI restriction sites (Figure 4.2). The resulting

fragment contained the entire coding region of rex, including a left flank of 636bp

and a right flank of 714bp. This fragment was then blunt end ligated into

pBlueScript. Inverse PCR (primers Grex_FLAG_F and Grex_FLAG_R) was used

to insert a 3XFLAG tag to the C-terminus of the rex gene immediately upstream of

the stop codon. The triple FLAG tag cassette, including the flanks, was isolated as a

BamHI fragment and cloned into the non-replicating conjugative vector pSX700.

pSX700::rex-3FLG was conjugated into into G. thermoglucosidans TM444, using

kanamycin resistance to select for first crossover homologous recombination events.

Screening for the loss of kanamycin resistance resulted in the isolation of a second

crossover 3xFLAG tagged rex allele which was confirmed via colony PCR (data not

shown). Initially experiments were performed to ensure 3xFLAG tagged Rex was

expressed. The FLAG tagged strain and WT cultures were grown in a variety of

aeration conditions and western blots performed. Oxygen limitation was created by

filling cultures to the top of tubes and closing the lid to restrict airflow. Western blot

results showed that 3xFLAG tagged Rex was expressed constitutively (Figure 4.3).

This expression pattern is consistent with B. subtilis tatAy and tatCy constitutive

expression, adding to the evidence that Rex in G. thermoglucosidans is part of the Tat

operon (Goosens et al., 2014).

99

Inverse PCR performed with primers including the FLAG tag sequence.

L_Flan rex R_Flan

The FLAG tagged rex gene was then cloned into a conjugative, non-replicating plasmid

pSX700

L_Flan rex FLAG R_Flan

BamHI BamHI

M N N E Q P K I P Q A T A K R L P L Y Y R F L

K N L H A S G K Q R V S S A E L S E A V K V D

S A T I R R D F S Y F G A L G K K G Y G Y N V

N Y L L S F F R K T L D Q D E I T E V A L F G

V G N L G T A F L N Y N F S K N N N T K I V I

A F D V D E E K I G K E V G G V P V Y H L D

E M E T R L H E G I P V A I L T V P A H V A Q

W I T D R L V Q K G I K G I L N F T P A R L N

V P K H I R V H H I D L A V E L Q S L V Y F L

K N Y P A E D Y K D H D G D Y K D H D I D Y K

D D D D K Stop

Figure 4.2 Diagram of the construction of a 3xFLAG tagged Rex allele. The in-frame FLAG tag

sequence is depicted.

100

4.2.2. Optimisation of chromatin immunoprecipitation

The Chromatin immunoprecipitation method described by Grainger et al.

(Efromovich et al., 2008) was optimised for G. thermoglucosidans (section 2.8).

Specifically, the cell disruption and sonication steps required improvement as DNA

fragments were not initially of the correct size and consistency for use in ChIP-seq.

The cell disruption step was first attempted with sonication (Bioruptor), then

optimised using cryogenic grinding, and the purified DNA fragments from each

method were analysed by gel electrophoresis (data not shown) for the amount of

desirable DNA of a consistent 250 bp size range. Cell disruption by sonication was

initially measured by a fall in OD600 measurement. Results indicated that cell

disruption was successful but fragment analysis revealed that although a strong

signal was detected, size was inconsistent. Cell disruption by cryogenic grinding

was far more successful and resulted in cleaner, more consistent fragment sizes. It

was reasoned that cryogenic grinding only disputed cells and had no impact on

DNA fragments, unlike sonication, so more consistent sizes resulted.

Aerobic (min) Oxygen limited (min)

0 0 30 60 90

100 110 120 130 140 150 160

WT 3xFLAG tagged Rex

Figure 4.3 Western blot of 3xFLAG tagged Rex being expressed under different oxygen abundant conditions over a time course.

101

Shearing of DNA was optimised by altering the number of biorupter cycles with the

aim of generating fragment sizes of between 100-500bp for maximum resolution. A

range of sonication cycles were performed on cryogenically lysed aliquots, and the

fragment sizes analysed on agarose gel. The optimal number of cycles using a

biorupter was found to be 35 x 30sec on/ 30 sec off which produced sizes of

between 100-600bp; this was considered acceptable.

4.2.3. Chromatin immunoprecipitation of G. thermoglucosidans

grown in liquid cultures

Biological replica cultures of strains TM444 rex-3xFLAG along the WT parental

strain (TM444) were grown in 50 ml TSB, 60oC with 250 rpm shaking. The WT

acted as a no antigen control, which provided an indicator of the amount of DNA

that is non-specifically purified during the assay. Strains were grown to late-

exponential phase (OD600 1.0-1.5) and then crosslinked by adding formaldehyde to

a final concentration of 1% to covalently stabilise the protein-DNA complex.

Cultures were incubated for a further 20 min and then any formaldehyde remaining

was quenched with glycine. The immunoprecipitation with anti-FLAG antibody

was performed resulting in an enriched DNA-protein sample. As a control, a

sample was taken prior to immunoprecipitation and referred to as total input DNA

control. These samples then underwent a series of wash steps to reduce non-specific

binding, followed by de-crosslinking and a DNA purification as described in section

2.8. The samples were purified using a Qiagen MinElute PCR purification kit, and

eluted in 22 µl of ultra-pure water.

4.2.4. ChIP-qPCR of predicted Rex target sites

The samples were analysed using quantitative real-time PCR (qPCR) to test

efficiency of crosslinking and immunoprecipitation specificity. Enrichment at

previously identified putative Rex target sites was measured using primers that bind

to the upstream regions of adhE and cydA genes, as well as a negative control GPI

primer set. Each set of data represents one biological repeat and a technical repeat

(Figure 4.4). The adhE was the greatest enriched by an average factor of 812, whilst

cydA was enriched by 202 fold, compared to the control GPI gene. The results

102

validated that enrichment occurred at putative Rex targets (adhE and cydA) and

samples were therefore analysed by deep sequencing.

4.2.5. Chip-Seq and identifying the Rex regulon

Samples followed the steps outlined in Figure 4.5 and were sent to The Genome

Analysis Centre (Norwich, UK) for library preparation and sequencing using the

Illumina HiSeq 2500 platform with a 50bp single-end read metric. Size selection

was 200-300bp and two biological repeats were run. Results were received as

FASTQ files and uploaded to Galaxy (http://usegalaxy.org) and converted to

Sanger, from Illumina format (Galaxy Tool version 1.0.4). Quality control was

performed using FASTQC (Galaxy Tool Version 0.63) and sequence data from the

two lanes on the HiSeq 2500 platform were concatenated. At the time of analysis

the genome most closely related to TM444 was that of Geobacillus thermoglucosidans

DSM-2542 and so the FASTQ files were then aligned with this genome using

Bowtie for Illumina (Galaxy Tool version 1.1.2) (Langmead, Trapnell, Pop, &

Salzberg, 2009). To avoid artefacts, the first 5 nt and last 5 nt of each read were

trimmed from each read prior to alignment and two mismatches were permitted.

adhE cydA gpi

Figure 4.4 Relative enrichment as determined by qPCR using immunoprecipitated DNA using

adhE, cydA and gpi primers. Values were then made relative for enrichment by subtracting

results of the no antigen control strain value from the 3xFLAG tagged Rex strain.

Rel

ati

ve

enri

chm

en

t

103

The output SAM files were compressed to BAM files using Sam to Bam (Galaxy

Tool version 2.0), and then visualised efficiently by creating bigWig files (Galaxy

Tool version 1.5.9.1.0) (Ramírez, Dündar, Diehl, Grüning, & Manke, 2014).

BigWig files were visualised using Integrated Genome Browser (IGB) to align the

histogram against the annotated Geobacillus thermoglucosidans genome. Rex

enrichment was broadly very specific, giving rise to some intense peaks across the

genome (Figure 4.6).

Figure 4.5 Flow chart of the stages involved in ChIP-seq data analysis.

Peaks were identified using MACS2 on Galaxy using a q value cut-off of <0.05.

This resulted in the overestimation of peaks (>4000 peaks) caused by the absence of

a control sample to calculate local background. Peaks with −log10 P-values of >20

were chosen for further analysis as listed in Table 4.2; peaks with lower −log10 P-

values were considered less likely to be biologically meaningful since many were

within open reading frames rather than regulatory regions. The final list consisted

of 22 genes and the corresponding Bacillus orthologues and reference have been

given alongside for comparison (Table 4.2). The peaks are illustrated by histograms

Sequencing, Illumina HiSeq 2500 platform.

FASTQ files uploaded to Galaxy.

Converted to Sanger, from Illumina format.

Quality control

performed using FASTQC.

FASTQ flies were then aligned to the

Geobacillus thermoglucosidans

genome.

FASTQ combined to a SAM format.

BAM files uploaded to Galaxy deepTools to produce bigWig file.

BigWig files visualised

using Intergrated Genome Browser (IGB)

104

in Figure 4.7 which illustrate the position of enrichment. These regions were then

investigated for Rex binding motifs and a G. thermoglucosidans consensus reached.

Many of the genes highlighted as part of the Rex regulon were previously identified

in other bacteria, however some new and unexpected discoveries were made.

Replica A

Replica B

Figure 4.6 Visualisation of BigWig histogram files with integrated Genome Browser. ChIP- seq analysis was performed on biological replicas of G. thermoglucosidans and the entire genome

represented.

105

Peak

name

Locus

tag

Function −log10

P-value

Distance

to start

codon

Location

consistent

with

regulation?

B. subtilis

orthologue

Reference to B. subtilis

MACS 2_peak

_32

AOT13_

00160

Geoth_0

611

NADH dehydrogenase

20.7602 -50 Yes ndh (YjlD) (Marino, Hoffmann,

Schmid, Möbitz, &

Jahn, 2000)

MACS

2_peak _846

AOT13_

03315

Geoth_3 897

Bifunctional

acetaldehyde- CoA/ alcohol

dehydrogenase

20.7602 -53 Yes adhA

adhB

(Romero, Merino,

Bolívar, Gosset, &

Martinez, 2007)

MACS 2_peak _1527

AOT13_

05975

Geoth_3 351

L-lactate

dehydrogenase 20.7602 -51 Yes ldh (Cruz Ramos et al.,

2000; Marino et al.,

2000)

MACS 2_peak _1818

AOT13_ 07200

Geoth_3 084

nitrite reductase

20.7602 -120 Yes narK (Cruz Ramos et al.,

1995)

MACS 2_peak _2340

AOT13_

09330

Geoth_2 650

polynucleotide

phosphorylase 20.7602 -175 Yes pnpA (Bechhofer & Wang,

1998; Farr,

Oussenko, &

Bechhofer, 1999)

MACS 2_peak _2739

AOT13_ 10935

Geoth_2

365

nitrite transporter

20.7602 0 Yes nirC (Reents et al., 2006)

MACS 2_peak _2689

AOT13_ 10655

hypothetical protein

20.7602 259 No - internal No homology

MACS 2_peak

_4364

AOT13_

18150 Geoth_0

897

6- phosphofructo

kinase

20.7602 -44 Yes pfkA (Tobisch, Zühlke,

Bernhardt, Stülke, &

Hecker, 1999)

MACS 2_peak _2256

AOT13_

08990

Geoth_2 719

signal

peptidase I 20.43195 -100 Yes sipS (Bolhuis et al., 1996)

MACS 2_peak _631

AOT13_ 02460

Geoth_0 118

50S ribosomal protein L1

20.43195 629 No - internal rplA

MACS 2_peak _1800

AOT13_

07120

fatty acid-

binding protein

DegV

20.43195 183 No - internal

MACS 2_peak

_3969

AOT13_

16625 Geoth_1

249

hypothetical

protein 20.43195 5 Yes, close to

N-terminus YqzE Functionally

uncharacterised

MACS 2_peak _1079

AOT13_

04165 alpha-amylase 20.1058 166 No - internal

MACS 2_peak _1816

AOT13_ 07190

Geoth_0

646

Nitrate/nitrite transporter

20.1058 -68 Yes NarK (Miethke, Schmidt, &

Marahiel, 2008)

MACS 2_peak

_2229

AOT13_

08865 Geoth_2

744

serine/threonin

e protein kinase

20.1058 1565 No - internal YabT (Bidnenko et al.,

2013; Pompeo,

Foulquier, &

Galinier, 2016)

MACS 2_peak

_2269

AOT13_

09045

recombinase

XerC 20.1058 448 No - internal

MACS 2_peak

_326

AOT13_ 01325

metallophosph oesterase

20.1058 1731 No - internal

106

MACS 2_peak _1782

AOT13_ 07050

Geoth_3 115

HxlR family transcriptional

regulator

20.1058 -358 Yes HxlR/yck

H

(Yurimoto et al.,

2005)

MACS 2_peak

_3135

AOT13_

12775 transposase 20.1058 468 No - internal

MACS 2_peak _3726

AOT13_ 15405

endonuclease III

20.1058 160 No - internal

MACS 2_peak

_4549

AOT13_

18745 transposase 20.1058 499 No - internal

MACS 2_peak

_909

AOT13_ 03545

spermidine synthase

19.78181 631 No - internal

Table 4.2 List of the enrichment sites identified by ChIP-seq. Included in the table is the gene identifier, locus tags, −log10 P-value, distance to the start codon, if enrichment is internal to a gene, and the homologous gene found in B. subtitles including references.

The Rex enriched regions in Table 4.2 are represented in histograms (Figures 4.7)

and can be segregated into three categories based upon location in relation to the

start codon, and whether or not the gene is predicted to be regulated by Rex based

on this position.

107

MACS2_peak_846

LuxR adhE

MACS2_peak_1527

drrABC ABC ldh lld

MACS2_peak_1818

SAM MFS HP narK

fnr

108

MACS2_peak_32

fpr

HP HP HP Nudi ndh

MACS2_peak_3969

HP

HP HP HP ComGF

MACS2_peak_4364

FxsA pyK pfkA AccA

MACS2_peak_1782

Hx1R

pgi 6PGD zwf

109

MACS2_peak_2340

Tru RibF rps pnpA

MACS2_peak_2739

PDH DLAT FNT HP

HP HP

MACS2_peak_2256

RimM trmD rplA sipS rsgA

MACS2_peak_1816

HP MobA SAM MFS HP

Figure 4.7 Histograms of Rex-3FLG enrichment sites and genes.

110

4.2.6. Putative Rex targets involved in fermentation and anaerobic

respiration

The strong MACS2_peak_32 is upstream from an NADH dehydrogenase (ndh)

gene, this enzyme is important for catalysing the transfer of electrons from NADH

to the quinone pool. The regulatory relationship between Rex and Ndh has been

referred to as a feedback loop that reduces oscillation of NADH/NAD+ (Gyan,

Shiohira, Sato, Takeuchi, & Sato, 2006). ndh is not likely to be part of a bigger

operon, however a divergent gene may also be controlled by the ROP site indicated

by the strong enrichment peak. The divergent gene is a ferredoxin-NADP reductase

(fpr), also transcribed alone, which has not been well studied in bacteria but may

play a role in adapting to oxidative stress and maintaining NADPH/NADP+

homeostasis (Jun Chen, Shen, Solem, & Jensen, 2015).

Enrichment at MACS2_peak_846 indicates that bifunctional acetaldehyde-CoA/

alcohol dehydrogenase (adhE) is Rex regulated. This gene is transcribed alone and

is integral to responding to oxygen limitation and a lack of a terminal electron

acceptor by fermenting to generate ATP.

L-lactate dehydrogenase (ldh) is a likely Rex target as enrichment upstream from

the start codon was found at MACS2_peak_1527. Downstream of the ldh gene is an

L-lactate permease encoding gene, involved in the transport of lactate across the cell

membrane, which is likely to occupy the same operon and is therefore also likely to

be regulated by Rex.

MACS2_peak_1816 and 1818 are located upstream of a predicted nitrate/nitrite

extrusion gene (narK) and a copper containing nitrite reductase (nirK) gene. It is

worth noting that the next gene downstream and convergent is an Fnr encoding

gene. In B. subtillis and E. coli both narK and Fnr form an operon (Cruz Ramos et al.,

1995). Analysis of this operon in B. subtillis demonstrated that a conserved Fnr

binding site was located upstream and Fnr was needed for induction and expression

of the operon (Reents et al., 2006).

Enrichment MACS2_peak_2739 indicated that nitrite transporter (nirC) is Rex

regulated. This gene is not part of a larger operon and is implicated in the nitrite

111

export system during anaerobic respiration to minimise toxicity (Ramírez et al.,

2000).

4.2.7. New Rex targets

Enrichment at MACS2_peak_4364 indicates that 6-phosphofructo kinase (pfkA),

the first gene of an operon including pyruvate kinase (pyK) is a Rex target. The pfkA

and pyK genes encode for enzymes that catalyse essentially irreversible reactions

and are therefore critical in controlling flux through the glycolytic pathway (Figure

4.8). PfkA catalyses the transfer of phosphate from ATP to fructose-6-phosphate to

form fructose-1,6-bisphosphate and ADP, whereas PyK catalyses the transfer of a

phosphoryl group of phosphoenolpyruvate (PEP) to ADP to form pyruvate and

ATP . Neither of these genes have previously been described as Rex targets,

although several other genes involved in glycolysis including PgK, Tpi, Fba have

been identified as part of the core Rex regulon (D. a Ravcheev et al., 2012). In B.

subtillis. pfkA and pyK also constitute an operon that is weakly induced by glucose

(Ludwig et al., 2001). The pfkA pyK operon has been studied in the bacterium

Corynebacterium glutamicum, an industrial amino acid producer, and it was

demonstrated that overexpression of glycolytic enzymes, including pfkA and pyK,

enhanced glucose consumption, improved alanine product yield and reduced

NADH/NAD+ ratio (Yamamoto et al., 2012). It seems that pfkA and pyK

upregulation during oxygen starvation or elevated NADH conditions may help to

produce ATP and rebalance NADH/NAD+ ratio by providing abundant substrates

for fermentation. Protein blast search revealed that 6-phosphofructo kinase (pfkA)

and pyruvate kinase (pyK) are the only copies of these genes within the G.

thermoglucosidans genome. It is implausible, therefore, that the regulation of this

operon is regulated simply by Rex, but it is expected that other regulators induce

expression and that Rex acts as a modulator.

Enrichment at MACS2_peak_1782 indicates that a ROP site is located upstream of

an HxlR family transcriptional regulator. In B. subtillis HxlR has been demonstrated

to be an activator of the formaldehyde inducing enzymes encoded by the hxlAB

operon involved in the ribulose monophosphate pathway (Yurimoto et al., 2005).

This operon was located downstream of HxlR in B. subtilis, and is similarly

positioned in G. thermoglucosidans genome, as just four genes separate HxlR and the

112

hxlAB operon. The hxlAB operon encodes 6-phospho 3-hexuloisomerase (hps) and 3-

hexulose-6-phosphate synthase (phi), used in formaldehyde fixation, which were

once thought to be confined to methylotrophs (Yasueda, Kawahara, & Sugimoto,

1999). An operon divergent to HxlR includes glucose-6-phosphate dehydrogenase

(Zwf) 6-phosphogluconate dehydrogenase (pgl) and glucose-6-phosphate isomerase

(pgi). However, our transcriptome analysis indicates (data not shown) this operon is

very unlikely to be a Rex target and in B. subtitles is confirmed as being constitutively

expressed (Ludwig et al., 2001).

Rex may control signal peptidase I (sipS) gene indicated by enrichment at

MACS2_peak_2256. Proteins such as signal peptidase I remove signal peptides

PFK

Fructose-6-P

PK

LDH Pyruvate Lactate

ADH

Ethanol

Glucose

Glucose-6-P G6DH

6-P-Gluconate

Figure 4.8 Brief outline diagram of the glycolytic and pentose phosphate pathway with important enzymes. Dotted arrows indicate that enzymatic reactions have been simplified and

omitted.

113

from secreted pre-proteins (Tuteja, 2005) and in B. subtillis sipS expression is

modulated to control protein secretion efficiency (Bolhuis et al., 1996). Another

enriched region was present upstream of a polynucleotide phosphorylase. This is an

enzyme found in eukaryotic, mitochondrial, chloroplasts and bacterial cells, and

functions to decay mRNA (Bechhofer & Wang, 1998; Farr et al., 1999).

The enrichment at MACS2_peak_2340 is located upstream of an operon encoding

a polynucleotide phosphorylase (pnpA) and a GTPase gene. PNPase is a 3’

exonuclease that decays mRNA, and although not essential in B. subtilis, ΔpnpA

strains demonstrate a clear phenotype including cells growing in chain formations

(B. Liu, Kearns, & Bechhofer, 2016).

Closer inspection of the ChIP-seq data to look at genes beyond the cut-off point of

−log10 P-value >20 revealed an enriched region upstream of fnr (Figure 4.9). fnr has

not been identified as a Rex target before but is known to be controled by the

ResDE system in many bacteria including B. subtilis (Geng, Zhu, Mullen, Zuber, &

Nakano, 2007).

Figure 4.9 Histograms of enrichment at fnr site and the surrounding region.

114

4.2.8. Rex binding site analysis

The sequence regions of Rex enrichment were analysed for a binding site motif

using http://meme-suite.org/, and the results listed in Table 4.3. Only peaks that

correspond with the start of a gene were included. Relative to the centre of each

peak, 50 bp upstream and downstream DNA was extracted for analysis. Using the

sequences of predicted ROP sites in Table 4.3, a potential consensus Rex binding

site was generated by compiling and aligning them and feeding them through a

sequence logo generator (Figure 4.10). The logo in Figure 4.11 indicates that ROP

sites may have a preference for GTG-n8-CAC, as do the majority of Rex ROP sites

across species (D. Ravcheev et al., 2012). The nucleotides residing within the centre

of the ROP site tend to favour adenine and thymine as do those eminently flanking

the GTG-n8-CAC motif.

http://meme-suite.org/

115

Peak number Target gene Position in

100 bp

around peak

P-value Binding site motif

MACS2_peak_

32

NADH dehydrogenase 70 1.02e-10 AACGTGAATATTTTCACAAA

MACS2_peak_

846

acetaldehyde dehydrogenase 67 5.53e-10 ATTGTGAAATGTTTCACAAA

MACS2_peak_

1818

nitrite reductase 38 7.22e-9 ATTGTTCATTATTTCACAAA

MACS2_peak_

1875

hypothetical protein 46 1.54e-8 ATCGTGAATAAAATCACAAA

MACS2_peak_

4364

6-phosphofructokinase 21 1.80e-8 TTCGTTCATTATTTCACAAA

MACS2_peak_

1527

L-lactate dehydrogenase 25 3.11e-8 ATTGTGCATTATTTCACAAT

MACS2_peak_

2340

polynucleotide

phosphorylase

34 5.27e-8 ATCGTGAATTGATTAACAAA

MACS2_peak_

2739

nitrite transporter 4 7.51e-8 AATGTGAAAATCATCACAAA

MACS2_peak_

2689

hypothetical protein 47 1.50e-7 AATGTGCATATGTTCGCGAA

MACS2_peak_

1816

MFS transporter 34 4.97e-7 TTTGTGACAAACATCACAAA

MACS2_peak_

991

stage II sporulation protein D 33 4.97e-7 ATCGCGCACTGTTTCACCAA

MACS2_peak_

3726

endonuclease III 54 5.45e-7 TTCGTTACTTTGTTCACCAA

MACS2_peak_

909

agmatinase 36 9.26e-7 AATCGGGAAAATTTCACGAA

MACS2_peak_

631

50S ribosomal protein L1 40 1.29e-6 ACGGTGACGTTTTTCACATA

MACS2_peak_

3969

YqzE family protein 22 1.40e-6 TACGTTAAGTTTTTCACGCA

MACS2_peak_

4195

pilus assembly protein PilM 50 1.40e-6 AACGTGAATATGTTAACATT

MACS2_peak_

2269

ATP-dependent protease

subunit HslV

51 1.64e-6 AATGTGACAACATTCACTAA

MACS2_peak_

4549

transposase 57 3.91e-6 TCCCGGCGAAGTTTCACGTA

MACS2_peak_

3135

transposase 26 3.91e-6 TCCCGGCGAAGTTTCACGTA

MACS2_peak_

1079

Hypothetical protein 38 5.11e-6 AATGTTTATTTAATCACGTA

MACS2_peak_

2256

signal peptidase I 40 6.57e-6 AACGTTCAATAATGAACATA

MACS2_peak_

2229

serine/threonine protein

kinase

77 3.04e-5 ATTCCGACAAAGTTCCGGAA

116

Table 4.10 Predicted ROP sites from each enriched peak region analysed by using a motif search

engine (http://motifsearch.com/index.php?page=motifseq). The start of the sequence in relation to the 100 bp sequence analysed is indicated.

4.3. Construction and analysis of a rex mutant in G.

thermoglucosidans

To further understand the role of Rex and to validate potential members of the Rex

regulon, a mutant strain was constructed and the expression of selected genes

analysed. The creation of a rex mutant strain was achieved by homologous

recombination, which involved transforming a rex mutant allele into G.

thermoglucosidans and applying selection to force the first homologous

recombination event, fusing the plasmid with the chromosome via integration at a

flanking region. The second recombination event was then screened for and

confirmed by colony PCR. Both transcriptome and metabolome data were collected

and analysed.

4.3.1. Initial attempts to isolate a Δrex mutant

An in-frame ∆rex mutant allele was constructed by M. Karigithu during his MSc,

which consisted of two 0.5kb flanking regions, which would produce a ∆rex mutant

protein with 175 central amino acids missing:

MSNEQPKIPQATAKR/GS/HHIDLAIELQSLVYFLRNYPLPS*. The mutant

allele was flanked by EcoRI sites and was initially cloned into the temperature-

Figure 4.11 Sequence logo derived from the genomic enriched regions

http://motifsearch.com/index.php?page=motifseq)

117

sensitive plasmid pTMO31. However, G. thermoglucosidans transformants could not

be isolated, possibly due to a problem with the kanamycin resistance gene on

temperature sensitive plasmid pTMO31.

The ∆rex allele was instead subcloned into an alternative temperature sensitive

plasmid pUB31 and used to transform G. thermoglucosidans by electroporation,

selecting for kanamycin resistance at 52 oC. Single cross-over recombinants were

then isolated at 68oC, purified via single colonies, then grown in absence of selection

to isolate second cross-over events (KanS). Colony PCR screening revealed several

potential mutants that were further tested for kanamycin sensitivity and further

rounds of colony PCR to ensure clonal populations. A ∆rex mutant could not be

isolated as only WT revertants and single crossovers could be identified.

An alternative approach was therefore taken by subcloning the ∆rex allele into a

non-replicative conjugative vector pSX700. However, although initial first crossover

recombinants were isolated, a second crossover ∆rex mutant was again not

obtained, possibly due to the low rate of recombination caused by the relatively

short flanking DNA. Furthermore, the difference in size of the flanks, 569 bp and

475 bp, would favour isolation of second crossover wild-type revertants. For this

reason, a new mutant allele was designed with longer flanks of similar size.

4.3.2. Recombination of Δrex allele into G. thermoglucosidans

New primers were designed (Section 1.2.3 F1MuantRex_6.12, R1MuantRex_6.12

and F2MuantRex_6.12, F2MuantRex_6.12) to amplify the Rex flanking region of

1336 bp and 1335 bp. The flanks were redesigned to delete 175 central amino acids

from the rex gene. The larger size exponentially increases the likelihood of

recombination events and the similarity in size, differing by only a single bp, results

in a theoretically equal likelihood of the second crossover event, increasing the

likelihood of isolating a rex mutant.

The flanking regions were amplified from chromosomal DNA, then independently

“blunt end” cloned into pBlueScript and the orientation and DNA sequence

confirmed. The upstream flank was then subcloned adjacent to the right flank using

BamHI and PstI restriction sites and the construct confirmed via restriction digest.

The resulting ∆rex allele was cloned into pSX700 via flanking EcoRI and PstI sites.

118

The vector pSX700-∆rex was transformed via conjugation and recombined into the

G.thermoglucosidans strain 11955 genome using kanamycin selection. Subsequent

removal of selection and screening for kanamycin sensitives allowed second

crossover strains to be isolated and the identification of a ∆rex strain. The ∆rex strain

was confirmed using colony PCR, which indicated that the wild-type allele was

replaced with a mutant allele with the loss of 175 amino acids.

4.4. Gene expression in a Δrex mutant

To confirm whether genes identified as putative Rex targets were indeed controlled

by Rex, RT-qPCR was performed using RNA samples isolated from the ∆rex strain

compared to those isolated from the WT parent. The impact of a variety of growth

conditions, specifically oxygen abundance, was also investigated.

4.4.1. Optimisation of growth conditions and RNA isolation

As there were no RNA harvesting methods available for G. thermoglucosidans, we

needed to design and optimise our own. There were two major steps to consider

during RNA isolation. Firstly, the lysis of cells without the degradation of RNA.

Secondly, the purification of RNA and degradation/removal of DNA. Initial

attempts using Qiagen kits failed to isolate intact RNA of a high enough

concentration. It was thought that this was due to poor cell breakage and therefore

a new method was designed that involved cryogenic grinding to disrupt cells. Cell

disruption was then followed by using an RNeasy midi kit (Qiagen) protocol. This

method yielded RNA concentrations 10 fold higher than the previous method (data

not shown).

Due to the large number of samples which were to be processed, the protocol was

further optimised. Cell disruption by sonication using a Biorupter replaced the

cryogenic grinding stage, thereby increasing maximum processing capacity from

two, to twelve samples per run. To ensure this method of cell disruption was

effective and did not degrade RNA, the samples were run on an RNA gel to check

ribosomal RNA was intact and abundant (data not shown). This method proved to

yield good quality RNA and was preferred.

119

For large-scale purification of RNA after cell disruption, the RNeasy midi kit was

replaced by phenol/chloroform extraction and isopropanol precipitation steps

described in Section 2.4.2.1. A DNase treatment kit (Ambion) was then used to

remove DNA. A reverse transcription kit (BioRad) was then used to generate

cDNA, and diluted 1/50 fold before use in qPCR. Primers for qRT-PCR were

designed using the Primer 3 tool (Section 2.2.3).

4.4.2. Transcript analysis of WT and Δrex mutant strains

The WT and Δrex G. thermoglucosidans strains were grown aerobically in 250 rpm

shaking flasks to OD600 1.0-1.5, at which point a 1.5 ml sample was taken (time

point = 0 min) and processed to isolate RNA. The remaining culture was used to

completely fill 1.5 ml microfuge tubes, and sealed such that no air gap remained.

The microfuge tubes were then incubated without shaking to produce oxygen

limited/fermentative growth conditions. At 30 min and 60 min intervals, the tubes

were processed to isolate RNA, and qRT-PCR was performed. The normalised

transcript abundance ratio was calculated by dividing Δrex by WT strain y-axis

values, and indicates the relative level of expression caused by the loss of Rex.

The transcript profiles for ldh, adhE, ndh and fnr genes are given in Figure 4.12.

Results here demonstrate that ldh is a likely Rex target as in the WT strain

expression increases during oxygen starvation at time point 60 min and the

expression profile in a rex mutant demonstrates deregulation at time point 0.

However, surprisingly ldh expression decreases in the Δrex strain following further

oxygen limitation, suggesting the involvement of additional effectors.

Unlike ldh, adhE expression remains supressed at T=0 in the Δrex strain and then, at

time points 30 and 60 min, is overexpressed in the Δrex strain compared to WT.

This again suggests that adhE is a direct target and that regulators other than Rex

may also be controlling adhE.

Expression of ndh differs from that of ldh than adhE. In WT ndh expression is

decreased by oxygen starvation. However, it is clearly a Rex target because the T=0

levels are ~4-fold increased in the Δrex strain. Nonetheless, the downregulaton of

ndh during oxygen limitation remains in the Δrex strain again suggesting that other

regulators may act to repress this gene.

120

The expression profile of fnr is of interest due to the role this regulator plays in

controlling anaerobic respiration. Furthermore, the similarity of Fnr and Rex DNA

binding site sequences (see Discussion), raises the possibility of interplay between

the two regulators. As would be expected, in the WT strain oxygen limited growth

conditions resulted in upregulation of fnr. In a rex mutant fnr was also induced

during oxygen limited growth conditions, but to a higher level, suggesting the

involvement of Rex in negatively regulating fnr expression.

Across all of the qRT-PCR experiments genes (ldh, ndh, adhE and fnr) in Δrex

strains, higher maximal expression levels were obtained demonstrating a partial loss

of control in responding to redox stress. However, no gene was constitutively

expressed, indicating that other regulators are likely to act on these genes. The

dissimilar nature of the ldh, ndh and adhE expression profiles indicates that the

regulators that act upon these genes might be different or might act differently due

to, for example, binding site location.

121

Figure 4.12 Absolute transcript of ldh, ndh, adhE and fnr measured by qRT-PCR. Error bars

indicate the standard deviation take of three biological repeats.

122

4.4.3. General growth characteristics of the ∆rex mutant in liquid

medium

Growth curves of both WT and ∆rex G. thermoglucosidans 11955 strains were

conducted to discover if the mutant strain grew differently, and to inform the design

of future fermentation experiments. Both WT and Δrex strains were grown first in

pre inoculum medium 2TY, and then in rich, buffered USM medium. Results

indicated that a loss of rex led to a reduced lag phase (Figure 4.13).

4.5. Fermentation analysis of Δrex mutant

To understand how the loss of rex might affect product formation, larger-scale

fermentation analysis was carried out. Strains were grown at our industrial

collaborators, ReBio Technologies, using specialist equipment to regulate oxygen

Figure 4.13 Wild type (TM960) and ∆ rex (TM961) growth curves in two types of media, 2TY

and USM. Experiments were conducted in shaking flasks and samples were taken at near hourly intervals.

Time (h)

Ab

sorb

an

ce (

OD

60

0)

123

exchange and measure substrate usage and product formation through HPLC. A

photograph of a fermentation tank used can be seen in Figure 4.14.

4.5.1. Growth and analysis of Δrex mutant in fermenters

To investigate the growth and product formation of G. thermoglucosidans Δrex and its

WT parent, strains were grown in 1L fermentation tanks with an aerobic to

anaerobic switch. This work was carried out collaboratively with ReBio

Technologies and the University of Sussex (specifically thanks to J. Neary, K.

Charman and Z. Walanus from ReBio Technologies). Samples were taken at

specific time points to analyse product formation and substrate utilisation.

Confluent TAS plates were harvested and used to inoculate seed flasks of 200ml

buffered 50 mM USM, 2% YE, 2% glucose at pH 7. These were grown to a target

OD600 of 8-10 before being used to inoculate the fermentation tanks. The 1L

Figure 4.14 Photographs of a 1L fermentation tank at ReBio Technologys. An additional

photograph of the tank from above with labels.

124

Time (h)

fermentation tanks were run in biological duplicate using media USM, 4% glucose,

2% YE, at pH 6.7. During the aerobic phase aeration was of the rate 1 Lpm until an

OD of 5-7 was reached, at which point conditions were switched to oxygen limited

conditions with an aeration rate of 0.2 Lpm. Samples were taken both before and

after the switch for metabolome (HLPC) analysis.

The growth curves WT 11955 and Δrex replicate strains are shown in Figure 4.15.

The replicate WT strains performed extremely similarly, however the Δrex strain did

not. One of the strains did not exhibit a lag phase and stopped growing early, at a

lower OD600 compared to the WT. The other strain grew similarly to the WT, but

reached a lower final OD600. Unfortunately we were unable to re-perform this

experiment due to limited time and resources (ReBio went into administration

shortly before the visit).

Figure 4.15 Growth curves of 11955 WT and Δrex strains in fermentation tanks. Medium was

USM, 4% glucose, 2% YE, at pH 6.7

Ab

sorb

an

ce (

OD

60

0)

125

4.5.2. HPLC analysis

HPLC was carried out on samples taken from the fermentation tanks. Both the

products produced, and the substrate glucose used during growth, were analysed for

changes in abundance (mM) periodically. HPLC analysis was performed by ReBio

technologies and the data sent to the University of Sussex for analysis. The

metabolic pathway in Figure 4.16 highlights the products being analysed, including

ethanol, acetate, succinate, citrate, lactate and some of the genes Rex is predicted to

regulate.

Metabolite analysis of the Δrex strain

The production of a variety of metabolites is represented in Figure 4.17 and 4.18.

The fermentative switch occurred at OD600 5 for all tanks and resulted in inducing

the production of formate, lactate and ethanol. The metabolites differ between the

Figure 4.16 The relevant metabolic pathways and products analysed by HPLC during

fermentation. LDH, lactate dehydrogenase; PFL, pyruvate formate lyase; PDH, pyruvate

dehydrogenase; AcDH, acetaldehyde dehydrogenase; ADH, alcohol dehydrogenase; PTA, phosphotransace- tylase; AK, acetate kinase; TCA cycle, tricarboxylic acid cycle. Adapted from (Cripps et al., 2009).

126

WT and Δrex strain primarily in ethanol and lactate yield. The Δrex strain has a

higher yield of lactate compared to the WT strain. This is consistent with the role of

Rex as a repressor of lactate dehydrogenase (ldh). Counter intuitively, however, the

Δrex strain produced less ethanol than the WT strain despite adh being a Rex target

gene.

127

Figure 4.17 Production of metabolites by G. thermoglucosidans WT and Δrex strains. The x-axis

represents the time post inoculation. Primary Y-axis represents glucose consumption (square)

and the of products ethanol (square), lactate (circle). Secondary y-axis represents the OD600 of the culture (dotted line and cross). The oxygen limiting switch occurred when the culture reached OD600 5 at which point aeration dropped from 1 Lpm and 600 RPM to 0.2 Lpm 300 RPM.

128

Figure 4.18 Production of metabolites by G. thermoglucosidans WT and Δrex strains. The x-axis

represents the time post inoculation. Primary Y-axis represents the concentration of products sucunate (circle), formate (square), citrate (diamond) and acetate (triangle). Secondary y-axis represents the OD600 of the culture (dotted line and cross). The oxygen limiting switch occurred when the culture reached OD600 5 at which point aeration dropped from 1 Lpm and 600 RPM to 0.2 Lpm 300 RPM.

129

The final yield of all metabolites was very similar in both strains, with the exception

of ethanol and lactate (see Table 4.3), which decreased and increased respectively.

Succinate yield was very slightly less in a Δrex strain, and acetate yield raised in one

of the two Δrex strains.

Product yield (mM product/mM glucose substrate used)

Ethanol

Lactate

Succinate

Formate

Citrate

Acetate

WT tank 2

0.31

1.37

0.04

0.08

0.02

0.13

WT tank 3 0.29 1.37 0.03 0.08 0.02 0.14

Δrex tank 4 0.13 1.63 0.02 0.08 0.02 0.14

Δrex tank 5 0.13 1.56 0.02 0.09 0.02 0.18

Table 4.3 Metabolite yields at point of full glucose substrate utilisation.

4.6. Future work and discussion

By combing genome wide Rex ChIP-seq data with analysis of Δrex strain via RT-

qPCR we confirmed putative Rex targets and uncovered potentially new ones.

Confirmation of the G. thermoglucosidans consensus sequence by analysis of Rex

enriched DNA regions was achieved. One example of a potential new target

unrelated to central metabolism and fermentation is the polynucleotide

phosphorylase (PNPase) encoding gene at peak 2340. Speculative answers as to

why this gene might be controlled by Rex could include a need for rapid turnover of

mRNA during times of changing redox state, so that previously expressed genes are

no longer translated and newly expressed genes are better represented. This would,

for example, allow the cell to more quickly adapt to oxygen limited conditions and

focus resources. The majority of genes potentially regulated by Rex included those

involved in fermentation, anaerobic respiration and, more generally, those involved

in the recycling of NADH to NAD+ and ATP production. However, some

enriched regions were internal to genes, suggesting that Rex plays a role in

controlling sRNA expression.

130

Fnr consensus sequence in

Bacillus subtilis

Rex consensus sequence in

Bacillus subtilis

4.6.1. Other regulators are likely to influence Rex target gene

expression

Transcriptional analysis of specific genes (Table 4.5) in a Δrex strain suggested the

involvement of additional regulators. One of these might be Fnr, which itself

appears to be regulated by Rex. Interestingly, the expression profiles of adh and ldh

differ, suggesting different mechanisms of control by Rex and/or additional

regulators.

For example, adh may have an additional repressor, or require an activator such as

Fnr, which lessens the effect of the Δrex mutation during oxygen abundant

conditions. Fnr binds in B. subtilis to ldh, inducing expression, and it could well be

true that Fnr regulates Rex targets in G. thermoglucosidans too (Reents et al., 2006).

It has been shown that in E. coli Fnr works in combination with other regulators

and that Fnr sites can be blocked by these regulators, preventing occupancy and

gene expression induction (Myers et al., 2013). Rex and Fnr binding sites are

strikingly similar (Figure 4.19) which could conceivably result in competition for

binding. In a Δrex strain Fnr might no longer have to compete with Rex, and so

may have increased access to its binding site to overexpress genes. The extent to

which Rex and Fnr compete may be dependent not only on the binding site but also

on oxygen and NAD+/NADH levels, the signals for activation of these regulators.

A system like this would help to differentiate between the need for expression of

anaerobic or fermentative genes.

Figure 4. 19 Consensus DNA binding site sequences of Rex in (Wang, Bauer, Rogstam, Linse, Logan, & von Wachenfeldt, 2008) and Fnr in Bacillus subtilis (Reents et al., 2006).

131

In the future, to investigate the possibility that Fnr and Rex are competing

regulators, an EMSA experiment could be carried out using Fnr and Rex proteins

with different binding sites and concentrations of NADH and O2. Chip-Sequencing

could also reveal the Fnr regulon and if Rex and Fnr sites overlap. Construction of

an fnr and rex double mutant and its use to investigate deregulation by qRT-PCR on

genes such as ldh and adh, would help to reveal if Fnr is necessary for expression of

these genes. Analysis of Fnr expression in a Δrex strain results showed that over

expression occurred during oxygen limited growth conditions only. Fnr expression

is controlled in B. subtilis by the two-component ResDE system, induced in response

to anaerobic conditions, but the exact signal is unknown (Geng, Zhu, et al., 2007;

Härtig & Jahn, 2012).

Gene

expression

profile

WT rex strain Δrex strain

O2

abundant

O2

limited

O2

abundant

O2

limited

Ldh expression Low Elevated Extreme Low

Adh expression Low Elevated Low Extreme

Ndh expression Elevated Low Extreme Low

Fnr expression Low Elevated Low Extreme

Table 4.4 Expression profiles found of ldh, adh, ndh, fnr in WT rex and Δrex stains in both oxygen

abundant and limited growth environments.

4.6.2. Fermentation and the metabolome

A Δrex strain was thought likely to increase the yield of fermentation products,

especially ethanol. Results demonstrate that lactate yield did increase in a Δrex

strain, but that crucially ethanol yield decreased. Although the reason behind this is

unknown, a probable cause is that the ADH enzyme was not the limiting factor,

and that LDH is a more efficient enzyme at pyruvate reduction and NAD+

regeneration. The deletion of LDH may result in ethanol yield increasing, although

such mutants in B. subtilis have a reduced growth rate. This may be because LDH is

able to use both NADH and NADPH and could be integral for equilibrium of

NADPH/NADP+ under fermentative conditions (Romero et al., 2007). For these

132

reasons future research into directing metabolic flux towards ethanol production

should consider the role of LDH in maintaining redox balance, substituting for

another enzyme capable of utilising NADPH, preferably an alcohol dehydrogenase.

An example of a strictly NADPH-dependent alcohol dehydrogenase was discovered

in a thermophilic Clostridium and converts acetoin to 2,3-butanediol (Ismaiel, Zhu,

Colby, & Chen, 1993; Korkhin et al., 1998). Although this particular alcohol

dehydrogenase is thermostable it does not produce ethanol and would not be

suitable, but it does indicate that potentially a NADPH-dependent one could be

discovered or even engineered to replace LDH.

133

5. Chapter V: Development of orthogonal redox -responsive

expression vectors

Overview

Bioinformatics analysis indicated that the Rex found in Clostridia might bind to an

alternative motif sequence compared to that of most other bacteria including

Firmicutes, Thermotogales, Actinobacteria, Chloroflexi, Deinococcis-Thermus and

Proteobacteria phylum. This raised the prospect of an orthogonal Rex

regulator/operator couple for Geobacillus, generating an alternative and

independently controlled regulatory network. This might, for example, allow target

genes to be induced at different redox potentials. The use of an orthogonal Rex,

with a differing redox sensing potential, may offer a means of optimising gene

expression for the production of ethanol or other fermentation products. To explore

this, Clostridium thermocellum Rex (C-Rex) was chosen for experimentation as this

bacterium grows at a similar temperature to G. thermoglucosidans. Electromobility

shift assays (EMSA) were used to confirm that C-Rex and G-Rex differ in their

DNA binding specificity at ROP sites. Expression vectors were then designed,

expressing synthetic FLAG-epitope-tagged Rex proteins, within a specially

engineered genetic circuit (several variations were constructed). To characterise the

functioning of the different Rex proteins, and the ability of the circuit to express

genes in a controlled manner, a fluorescent reporter gene was put under the control

of a variety of ROP sites. qRT-PCR and western blot analysis were used to

investigate RNA expression and protein levels, in response to changing

environmental oxygen abundance, and to assess the success of the expression

circuits.

5.1. Identification of potentially orthogonal and thermostable Rex

regulators

Clostridia have been proposed to possess a different Rex DNA-binding motif and

Rex protein compared to most other Gram-positive bacteria (see Figure 5.1) (D. A.

Ravcheev et al., 2011; D. Ravcheev et al., 2012). Experiments have been conducted

that determine if the different binding sites can be used as an orthogonal Rex

134

regulatory system with the ability to regulate fermentative genes independently of

the host Rex. A search for Rex proteins in the thermophile C. thermocellum found

that it possesses two rex paralogs, Cthe_0422 and Cthe_1798. Although it is rare for

a bacterium to possess rex paralogs, C. thermocellum is not alone. Clostridium bartlettii

and Thermotogales maritima also have paralogues and are some of the most

divergent, with the largest number of Rex amino acid substitutions being found in

T. maritima (D. a Ravcheev et al., 2012). It is not currently known why some

bacteria have paralogues or even whether both paralogues sense NAD(H).

Cthe_0422 appears to be cotranscribed with bi-functional acetaldehyde-

CoA/alcohol dehydrogenase (adhE), whereas Cthe_1798 appears to be

monocistronic, and surrounded by genes unrelated to fermentation or central

metabolism. The C .thermocellum paralogues were aligned with the Rex protein

sequences from G. thermoglucosidans and B. subtilis (Figure 5.2). Rex Cthe_1798

shared a 38% identity and Cthe_0422 shared a 35% identity to B-Rex. G-Rex

shared an 83% identity with B-Rex. The protein alignment includes the Rossmann

fold (NADH binding domain) and the winged helix-turn-helix (DNA binding

domain). Of particular interest are the 10 residues, Pro4, Arg10, Ser31, Arg46,

Lys47, Tyr51, Gly56, Gly59, Gly61, Tyr62, which are highly conserved throughout

the Rex family and T-Rex solved structure indicates they contact DNA

(McLaughlin et al., 2010). Phe43 and Arg58 are also important for Rex-DNA

interaction, however these are far less well conserved. Cthe_0422 protein sequence

has several substitutions of key residues including 4, 31, 43, 51, 58 and 59 and the

protein sequence of Cthe_1798 also contains slightly fewer substitutions. Residue

47 is of particular importance as it contacts and forms a hydrogen bond with the 5th

C .thermocellum - TTGTTAANNNNTTAACAA

G. thermoglucosidasius – TTGTGAANNNNTTCACAA

Figure 5.1 Proposed Rex binding site of C .thermocellum compared to that of G.

thermoglucosidasius. Underlined are the key differing nucleotides.

135

T-Rex 1 -----MKVPSAAISRLITYLRILEELEAKGIHRTSSEQLAEEAQVTAFQV

G-Rex 1 MNNEQPKIPQATAKRLPLYYRFLKNLHASGKQRVSSAELSEAVKVDSATI

Cthe_0422-Rex 1 -MNNPKAISKQTLLRLPSYLSYLKSLPKQDGEYVSATMIASALGLNDVQV

Cthe_1798-Rex 1 -MNLDKKISMAVIRRLPRYYRYLSDLLKLGITRISSKELSSRMGITASQI

B-Rex 1 MNKDQSKIPQATAKRLPLYYRFLKNLHASGKQRVSSAELSDAVKVDSATI

consensus 1 . ... ....**. *...*. * . . ..*. ...... . ..

T-Rex 46 RKDLSYFGSYGTRGVGYTVPVLKRELRHILGLNRRWGLAIVGMGRLGSAL

G-Rex 51 RRDFSYFGALGKKGYGYNVNYLLSFFRKTLDQDEITEVALFGVGNLGTAF

Cthe_0422-Rex 50 RKDLACVSKKGRPKLGYVAEELIRDIESFLGYDSLNDAIIVGAGRLGGAL

Cthe_1798-Rex 50 RQDLNCFGGFGQQGYGYNVEYLYKEIGNILGVNEAFKIIIIGAGNMGQAL

B-Rex 51 RRDFSYFGALGKKGYGYNVDYLLSFFRKTLDQDEMTDVILIGVGNLGTAF

consensus 51 *.*...... *....**....*...... *. ... .....*.*..*.*.

T-Rex

96

ADYPGFGETFELRGFFDVDPEKIGKPVGRGVIEPMEALPQRVPG-RIEIA

G-Rex 101 LNYNFSKNNNTKIVIAFDVDEEKIGKEVGGVPVYHLDEMETRLHE-GIPV

Cthe_0422-Rex 100 LSYEGFKEYGLNIVAAFDIDESKIGTEICGRKIFPLDKMKELCRRMKIRI

Cthe_1798-Rex 100 ANYTNFEKRGFKLIGIFDINPNLIGKKIRDVEIMHLDSLDRFVAENQVDI

B-Rex 101 LHYNFTKNNNTKISMAFDINESKIGTEVGGVPVYNLDDLEQHVKD--ESV

consensus 101 . * .. ... .... . ... .. .. .. .. .. ... . .

T-Rex 145 LLTVPREAAQEAADQLVRAGIKGILNFAPVVLEVPKEVAVENVDFLAGLT

G-Rex 150 AILTVPAHVAQWITDRLVQKGIKGILNFTPARLNVPKHIRVHHIDLAVEL

Cthe_0422-Rex 150 GIITVPADSAQKVCDMLVDSGIYAIWNFAPVHLKVPDNILVQQENMAASL

Cthe_1798-Rex 150 AILCVPYENTPAVADKVARLGVKGLWNFSPMDLKLPYDVIIENVHLSDSL

B-Rex 149 AILTVPAVAAQSITDRLVALGIKGILNFTPARLNVPEHIRIHHIDLAVEL

consensus 151 ....... .. . .... ..... ....... ... ... . .. .

T-Rex 195 RLAFFILN--------------------

G-Rex 200 QSLVYFLKN-------------------

Cthe_0422-Rex 200 AVLSAHLAEAIKNNGIGKGDEENDENSE

Cthe_1798-Rex 200 MVLGYRLNEMRKSQRNKS----------

B-Rex 199 QSLVYFLKHYSVLEEIE-----------

consensus 201 .. . *

G/T nucleotide of Rex operator sites and inserts into the major groove of DNA,

nucleotide 5 is contacted by Lys47 in Rex proteins from Actinobacteria and

Thermus, or Arg47 in Bacilli Rex, both of which are positively charged amino acids

(L. Zhang et al., 2014). The Cthe_0422 protein also contains a Lys47 residue,

however in Cthe_1798 protein it has been substituted for Gln (Luscombe,

Laskowski, & Thornton, 2001). The Gln47 has only been found in Clostridia and

correlates with the Rex DNA binding motif: TTGTTAANNNNTTAACAA as

opposed to the binding site found both in Actinobacteria and Bacilli:

Figure 5.2 Sequence alignment of T. aquatics, C .thermocellum, G. thermoglucosidans and B.

subtilis Rex homologues. Black highlights residues that are the same, grey those with similar

properties and white which are markedly different.

136

TTGTGAANNNNTTCACAA (D. A. Ravcheev et al., 2012). The key differences

being at positions G5 and C14, changing to T5 and A14. This is likely to result in

alteration of the way Rex can interact with DNA. To understand more about the

activity of C. thermocellum Cthe_0422 and Cthe_1798 Rex and how they interact

with ROP sites compared to G. thermoglucosidans Rex, the proteins were expressed,

purified and used in EMSA experiments to assess their binding affinity for a variety

of operator sites.

5.2. Cloning and overexpression of Clostridium thermocellum rex

paralogues

Both C. thermocellum Rex paralogs, along with the G. thermoglucosidasius Rex were

over expressed and purified in preparation for EMSA experiments to investigate

relative affinities for the artificial “ideal” ROP sites: TTGTTAANNNNTTAACAA

and TTGTGAANNNNTTCACAA were designed as ideal sites for clostridial Rex

and Geobacillus Rex, respectively.

5.2.1. Design and overexpression of rex

C. thermocellum Rex (Cthe_0422 and Cthe_1798) were codon-optimised for E. coli

expression and the sequences ordered from Invitrogen as double stranded linear

DNA Strings™ Fragments with flanking NdeI and BamHI restriction sites. The

two genes, along with a previously constructed G-rex NdeI-BamHI cassette

(Michael Murage Karigithu, 2012) were cloned independently into NdeI / BamHI-

digested pET15b which fused, then overexpressed using E. coli strain

BL21λDE3(pLysS) (Section 2.5.1). Rex was purified by Ni-affinity chromatography

(Section 2.5.2), then buffer exchanged using VIVASPIN-500 (section 2.5.3) to

lower imidazole and salt concentrations and to concentrate the protein to 150-950

µg/ml. The proteins were stored in Storage buffer, on ice at 4 oC overnight. For

longer storage periods proteins were suspended in 40% glycerol, snap frozen using

liquid nitrogen and stored at -80 oC until needed.

While C. thermocellum Rex paralogues (Cthe_0422 and Cthe_1798) were

successfully expressed and purified, G-Rex was found to be quite insoluble, and

was only present at low levels in the soluble cell lysate. Multiple 50 ml small-scale

137

methods were trialled and optimised before scaling up to 500ml for a larger yield. A

reduced rate of protein synthesis, by a combination of a lower growth temperature

and a lesser concentration of IPTG (Table 5.1), was found to be successful for

helping to increase the solubility of G-Rex. The slower rate of protein synthesis

allows time for chaperones and foldases to be expressed at the proteins to fold

correctly, promoting solubility (Sørensen & Mortensen, 2005). Figure 5.3

demonstrates the difference in solubility when reducing the rate of protein

synthesis. G-Rex is more soluble when expressed under optimised conditions and

became more concentrated in the cleared cell lysate (lane 4), compared to the whole

cell lysate (lane 5).

Culture trial OD600 prior to induction

Initial growth temp

IPTG concentration

(mM)

After induction

growth temp

Growth duration (hour)

Original 0.4 - 0.7 37 oC 1 30 oC 3.0 Final

optimised 0.3 - 0.4 37 oC 0.4 16 oC 16

Table 5.1 Growth conditions of original methods of overexpressing G-Rex, and final

optimised method. Changes in conditions aimed to slow the rate of expression and help

promote folding, to produce more soluble protein.

138

5.3. Design of EMSA probes

Short (~24 bp) DNA oligonucleotide probes were designed based upon predicted

Rex targets, following RSAT (http://embnet.ccg.unam.mx/rsa-tools/) genome-

wide searches using the consensus sequence inputs: TTGTTAANNNNTTAACAA

for C. thermocellum (DSM 1313); and TTGTGAANNNNTTCACAA for G.

thermoglucosidans (C56-YS93). C. thermocellum results included the nuoA-N operon

with sequence TTGTTAAATTGTTAACAT, pensioned -49 in relation to the start

codon. An ROP site upstream of the nouA-N operon in Streptomyces coelicolor has

also been identified and in vitro binding assays confirmed Rex bound (Brekasis &

[kDa] 18

38

28

G-Rex monomer

(24 kDa)

Figure 5.3 Optimisation of G-Rex expression. Samples were assessed by separating proteins using 12% SDS PAGE. Numbers to the left represent molecular weight maker (M) in kDa. Lanes 1-2 represent samples taken during the original method of G-Rex protein expression. Lane 1 is the cleared cell lysate, lane 2 is the whole cell lysate. A clear loss of G-Rex in lane 1 (compared to lane 2) was caused by the insolubility of the protein and it pelleting during centrifugation. Lanes 3-4 represent samples taken during the final optimised method. Lane 3 is

the cleared cell lysate, lane 4 is the whole cell lysate. Instead of a loss of G-Rex, it becomes more concentrated in lane 4 (compared to lane 5), due to the removal of the cell debris,

indicating it is far more soluble when expressed under the optimised conditions.

http://embnet.ccg.unam.mx/rsa-tools/)

139

Paget, 2003) and the ndh gene in Bacillus subtilis has been confirmed as a Rex target

in vitro (Gyan et al., 2006). Due to this the nuoA-N operon was deemed to be a

likely Rex target and a probe was designed (named CT1T, table 5.1). A second

potential Rex binding site in C. thermocellum site was found manually. It was

thought likely that the C-Rex paralogue, which is located upstream of an adhE gene

located within an operon, is a Rex target. A site resembling a ROP was located —

211 in relation to the C-Rex paralogue start codon, a probe was also designed based

upon this sequence (named C0422, Table 5.2). Two probes were also designed

based upon G. thermoglucosidasius Rex consensus sequence. The predicted Rex

binding site within the promoter region of ldh (lactate dehydrogenase) was used to

design a probe GTLD1. The other G. thermoglucosidasius probe was based on the C-

Rex probe CT1T, but the key nucleotides (at positions 5,14) were substituted to

switch specificity to match the G. thermoglucosidans ROP consensus.

The sequences were ordered as single stranded DNA oligos from Eurofins and one

of the oligos (per probe) was end-labelled with ϒ-ATP. The probes were then

annealed using an excess of the unlabelled oligo. These were purified on a NAP-5

column and eluted in TE buffer ready for use during EMSA. During EMSA

experiments probes were mixed with Rex proteins, incubated for 15 min at room

temperature, and analysed non-denaturing 6% polyacrylamide gels.

Name ROP short hand ROP Sequence

CT1T ROPT5/A14 TTGTTAAATTGTTAACAT

CT1G ROPG5/C14 TTGTGAAATTGTTCACAT

GTLD ROPG5/C14 TTGTGAAATAATGCACAA

C0422 ROPG5/A14 TTGTGAAAACTATATAAA

Table 5.2 EMSA probe Rex binding sites. Comparison and names of the Rex binding sites chosen for EMSA experiments. Key nucleoties are highlighted in bold.

140

5.4. DNA binding analysis of Rex proteins using EMSA

Initially the interaction of G-Rex with CT1G and CT1T probes that differ by the

presence of G or T at position 5 (and corresponding change of C to A at position

14) in the ROP sequence was investigated. As predicted, G-Rex bound more

strongly to CT1G compared to CT1T (figure 5.4). This confirms the importance of

G5/C14 for G. thermoglucosidasius Rex –DNA interactions and suggests that ROP

sites that contain T5/A14 might be poorly recognised by G-Rex in vivo. In addition

to Rex-DNA complexes that migrated just above the free probe, a smear of slower-

migrating higher order complexes was also present.

EMSA studies with C-Rex Cthe_1798 and Cthe_0422 paralogues demonstrated

that Cthe_1798 showed strongest binding to the CT1T probe, with only weak

binding to CT1G or to GTLD1 (Figure 5.5). Surprisingly, the paralogue Cthe_0422

did not bind to any of the probes (CT1T, CT1G, GTLD1). A fourth probe named

C0422 was designed (as previously described) based on the predicted ROP site

upstream of Cthe_0422 and the nouA-N operon. However Cthe_0422 Rex also

failed to bind to this probe (data not shown). This result was unexpected as

Cthe_0422 shares its operon with bi-functional acetaldehyde-CoA/alcohol

dehydrogenase, often a Rex target. In light of this, experiments using Cthe_0422

140

were discontinued and remaining studies focused on Cthe_1798, which is referred

to as C-Rex.

CT1T (ROPT5/A14) CT1G (ROPG5/C14)

Figure 5.4 G-Rex binds to the G. thermoglucosidasius consensus sequence probe more strongly than to the probe comprising of a C. thermocellum consensus sequence. EMSA binding reactions contained 1ng of CT1T or CT1G 5’ end labelled DNA probe, 0.2 µg of non- homologous herring sperm DNA, increasing concentrations of G-Rex protein. Black arrows

indicate G-Rex DNA complexes. Grey dashed arrow indicates unbound probe.

G-Rex bound probe

Free probe

G-Rex G-Rex

P P

141

P

Free probe

C-rex bound

probe

5.4.1. C-Rex is responsive to NADH and NAD+

To investigate if the DNA binding activity of C-Rex is responsive to pyridine

nucleotides, EMSA experiments were performed in the presence of NADH and

NAD+. Figure 5.6 demonstrates that as concentrations of NADH increase, DNA-

binding decreases.

P P

CT1T

(ROPT5/A14)

CT1G

(ROPG5/C14) GTLD1

(ROPG5/C14)

Figure 5.5 C-Rex Cthe_1798 binds to the C. thermocellum consensus sequence probe more

strongly than to the probe comprising of a G. thermoglucosidasius consensus sequence. EMSA binding reactions contained 2 ng of CT1T, CT1G or GTLD1 5’ end labelled DNA probe, 0.2µg of non-homologous herring sperm DNA, 0.1 µM of either Cthe_0422 or Cthe_1798 proteins. Black arrows indicate C-Rex DNA complexes. Grey dashed arrow indicates unbound probe.

142

To determine whether C-Rex was also responsive to NAD+, experiments were

repeated in the presence of 1 mM NAD+, with C-Rex of increasing concentrations

(Figure 5.7). NAD+ appeared to improve DNA-binding. This is not the case

throughout the Rex family. NAD+ does not increase S. coelicolor Rex binding

affinity, however it does in the case of the Rex orthologues from S. aureus and

importantly B. subtilis (Brekasis & Paget, 2003; Gyan et al., 2006; Pagels et al.,

2010).

The EMSA data indicate that the C. thermocellum Rex paralogue Cthe_1798 possess

qualities such as an alternative consensus binding sequence (compared to G-Rex)

and redox sensing, which makes it a suitable choice in the design and construction

of Rex-based orthogonal expression vectors.

Figure 5.6 C-Rex DNA binding activity is inhibited by increasing concentrations of NADH.

EMSA binding reactions contained 1 ng of CT1T 5’ end labelled DNA probe, 0.2 µg of non- homologous herring sperm DNA, C-Rex protein, increasing concentrations of NADH from 0 mM, 0.001 mM, 0.01 mM, 0.1 mM, 1 mM. Black arrow indicates C-Rex DNA complex. Grey dashed arrow indicates unbound probe.

Free probe

C-rex bound

probe

NADH

P C-Rex

CT1T

143

5.5. Design and construction of Rex-based genetic circuit

expression vectors

To investigate the usefulness of a C-Rex-based expression system in vivo, a set of

plasmids containing modular components was assembled. A variety of promoters

were designed that included ROP sites based on either C-Rex (ROPT5/A14) or G-Rex

(ROPG5/C14) consensus sequences (labelled in Figure 5.8 as Rex operator), in a single

or double ROP site format. These promoters were located upstream of a reporter

gene gfp. Plasmids also contained a copy of either C-Rex or G-Rex genes, the

expression of which was auto-regulated through the inclusion of another ROP site.

Expression studies were carried out in the G. thermoglucosidasius 11955 Δrex strain to

prevent interference from the host Rex. Initially Western blot analysis was

performed to investigate whether the C-Rex and G-Rex proteins were expressed and

controlled in a redox dependent manner. Their expression levels were compared to

a G. thermoglucosidasius 11955 3xFLAG tagged Rex strain. Expression of the gfp

reporter gene and that of selected genomic rex targets was also then investigated by

qRT-PCR. Figure 5.6 is a schematic of the genetic circuit and

NAD+

Figure 5.7 C-Rex DNA binding activity is enhanced by NAD+. EMSA binding reactions contained 2 ng of CT1T 5’ end labelled DNA probe, 0.2 µg of non-homologous herring sperm DNA, 1 mM NAD+, increasing concentrations of C-Rex protein. Black arrows indicate Cthe_1798 C-Rex DNA complexes. Grey dashed arrow indicates unbound probe.

C-Rex C-Rex

P P

CT1T

Higher order

complexes

C-rex bound

probe

Free probe

144

illustrates that Rex expression leads to negative regulation of gfp and autoregulation

of rex itself.

5.5.1. Design of rex gene modular components and auto-regulating

promoter

The design of the C-Rex gene modular fragment was based on the C. thermocellum

cthe_1798 gene, which was found to be the only of the two paralogs which bound

to probes during EMSA experiments. This gene was codon optimised for Bacillus

subtillus, as was the G-Rex gene, to help ensure both would be expressed in a very

similar way in G. thermoglucosidans. A 3xFLAG tag was placed just before the stop

codon of rex (see in yellow Figure 5.9) for western blot analysis of protein levels.

Streptococcus mutans and S. coelicolor have been found to express Rex via an auto-

regulated feedback loop (Bitoun, Liao, Yao, Xie, & Wen, 2012; Brekasis & Paget,

2003), and bioinformatics has predicted that out of a total of 119 genomes

analysed, 51 demonstrate Rex autoregulation (D. A. Ravcheev et al., 2012). An

auto-regulatory mechanism may be preferable for this project so rex is not

overexpressed on multicopy plasmids. Auto regulation was incorporated into the

design as it would also prevent overexpression that could occur when using a

constitutive promoter. Negative autoregulation has also been found to increase

Figure 5.8 Schematic of the rex genetic circuit module.

145

response time, leading to a more dynamic gene expression system (Rosenfeld,

Elowitz, & Alon, 2002). Therefore to control Rex expression, modular fragments

were designed to include a promoter sequence that included a ROP site. This

promoter was based on the pUP, which derives from the constitutively expressed

uracil phosphoribosyl transferase gene of G. thermoglucosidans and is functional in

both E. coli and G. thermoglucosidans (Reeve et al., 2016). From the pUP promoter

library, pUP3 was selected as it possessed one of the midrange expression profiles.

The promoter was modified so that a Rex binding site was placed partially

overlapping the -35 region (Figure 5.9). As this results in a suboptimal -35 region,

an extended -10 region was also added as this element enhances Bacillus promoter

strength, and was found to mitigate the loss of similarity within the -35 region in E.

coli (Voskuil, Voepel, & Chambliss, 1995). A TRTG motif was found to be

conserved in extended -10 regions of Bacillus sp. (G. Chen, Kumar, Wyman, &

Moran, 2006). The sequence TGTG was chosen as crucially only a single

nucleotide of the pUP3 sequence would need to be substituted to achieve it (figure

5.10).

The G-Rex and C-Rex promoters were identical, except for the two discriminatory

nucleotides within the ROP site. An XhoI restriction site (CTCGAG) was placed

CGGCGATTAT AAGGACCACG ATATCGACTA CAAAGACGAT

1 ATGCCTCGAG GTCAATGATT GTTAAATACT TAACAATGTG ATAGAATGGT GTTTAGGAAA AGCGTACGCC GTAACAAAGA

81 GGAGAGTGCA AAATGAACCT GGACAAAAAA ATCAGCATGG CAGTTATTCG TCGTCTGCCT CGTTATTATC GTTATCTGAG

161 CGATCTGCTG AAACTGGGTA TTACCCGTAT TAGCAGCAAA GAACTGAGCA GCCGTATGGG CATTACCGCA AGCCAGATTC

241 GTCAGGATCT GAATTGTTTT GGTGGTTTTG GTCAGCAGGG TTATGGTTAT AATGTGGAAT ATCTGTACAA AGAGATCGGC

321 AATATTCTGG GTGTTAATGA GGCCTTCAAA ATCATTATTA TCGGTGCCGG TAATATGGGT CAGGCACTGG CAAATTATAC

401 CAACTTTGAA AAACGCGGTT TCAAACTGAT CGGCATCTTT GACATTAATC CGAACCTGAT CGGGAAAAAA ATCCGTGATG

481 TTGAAATCAT GCATCTGGAT AGCCTGGATC GTTTTGTTGC AGAAAATCAG GTTGATATTG CCATTCTGTG TGTGCCGTAT

561 GAAAATACAC CGGCAGTTGC AGATAAAGTT GCACGTCTGG GTGTGAAAGG TCTGTGGAAT TTTAGCCCGA TGGATCTGAA

641 ACTGCCTTAT GATGTGATTA TCGAAAATGT GCATCTGTCC GATAGCCTGA TGGTTCTGGG TTATCGTCTG AATGAAATGC

721 GTAAAAGCCA GCGCAATAAA AGCGACTACA AAGATCATGA

801 GACGATAAGT AATTAATTAA ATGC

Figure 5.9 Synthesised C-Rex fragment modular component. Sequence includes a specially designed auto regulating promoter, and a FLAG tag. Pink highlights are Rex binding sites. Green highlights represent the extended -10 region. Underlined font colour indicates -35 or -10

region. XhoI, BsiWI and PacI restriction sites are highlighted in blue. Yellow highlighted nucleotides are the 3xFLAG tag.

146

upstream the promoter and a BsiWI (CGTACG) site downstream, just before the

ribosome binding site (RBS). These sites allow for exchange of components so that

other promoters could be studied in the future. The C-Rex and G-Rex modular

components, including the auto-regulating promoters, were synthesised by

Invitrogen as gene Art double stranded strings, and cloned into pG1AK-

sfGFP_OriT via XhoI and PacI sites (Figure 5.11).

Upstream element -35 -10 RBS sequence Start codon

pUPWT

pUP3

adh promoter

TAAAGATTTGTTTTAATACAACTTGACACTTTGTGTAGAAT

Auto G-Rex

EMSA CT1G

CTAACTGTTGTGAAATTGTCACAACCTAACTGC

Auto C-Rex

EMSA CT1T

CTAACTGTTGTTAAATTGTAACAACCTAACTGC

Figure 5.10 The design of Auto G-Rex promoter. Sequence of pUPWT a uracil phosphoribosyl transferase gene promoter in G. thermoglucosidans. pUP3 from a promoter library based on pUPWT. adh promoter region, including an extended -10 region. Auto G-Rex promoter

designed based on pUP3 and adh extended -10 region, incorporating a G-rex binding site. The EMSA CT1G probe sequence including a Rex consensus binding site. Auto C-Rex promoter. EMSA CT1T probe sequence. Hilighted in blue are XhoI and BsiWI restriction sites. Pink highlights are Rex binding sites. Green highlights represent the extended -10 region. Red font colour indicates -35 or -10 region.

147

XhoI

NotI

5.5.2. Analysis of the auto regulating promoter controlling rex gene

expression

It was expected that the plasmid-borne rex genes should be autoregulated such that

a shift in the NADH/NAD+ redox poise would increase the cellular level of Rex

protein. To test this, western blots were performed, and the data compared to that

obtained with G. thermoglucosidasius 11955 3xFLAG tagged rex strain, in its natural

locus. Overnight TSA plate cultures were used to inoculate 50 ml TSB, then grown

aerobically with shaking to an OD600 of 1.5. Following harvesting of an aerobic

sample, two 15 ml tubes were filled completely, sealed to eliminate air, then

incubated without shaking for 30 or 60 mins, prior to harvesting. Following the

BsiWI XbaI

PacI SacI

p11AK2_OriT_Re

PmeI

FseI

EagI

Figure 5.11 Plasmid Map of p11AK2_OriT_Rex.

148

washing of pellets 0.9% NaCl (w/v) they were resuspended 1 ml of ice-cold Lysis

buffer (0.9% NaCl, containing protease inhibitors). 250 µl of 50% (v/v) TCA was

then added and, following incubation on ice for 15 min, samples were fully

disrupted using a biorupter. The TCA insoluble material, including protein, was

harvested by centrifugation, washed with 100% acetone, then finally resuspended in

a denaturing SDS-PAGE gel loading dye prior to SDS PAGE and Western

analysis.

The results revealed that Rex expressed from the native locus is constitutively

expressed under these growth conditions in the WT strain. This concurs with the

absence of an obvious Rex binding site upstream of the operon and the absence of a

Rex peak in the Chip-seq experiments. It therefore appears that Rex is not

autoregulated or even induced by oxygen limitation through a different mechanism

in Geobacillus. However, as expected, when C-rex or G-rex were placed under the

control of promoters that included ROP operators, protein levels increased during

oxygen limitation (Figure 5.12). It therefore appears that each Rex protein is active

for both DNA binding and NAD(H) sensing in vivo: C-Rex and G-Rex bind to

their respective ROP sites to inhibit expression during high oxygen abundance and

low NADH levels, then dissociate when aerobic respiration is inhibited and NADH

levels rise. Presumably, basal levels of Rex ensure very low levels of expression

under oxygen abundant conditions.

149

G - Rex Genomic Rex C - Rex

Importantly, plasmid-expressed C-Rex and G-Rex protein levels did not exceed that

seen for the WT strain, which should minimize the possible pleiotropic effects that

might result from excess regulator. These data indicate that the auto-regulating

promoter worked as intended and that experiments should proceed.

5.5.3. Design and analysis of promoter controlling gfp gene

expression

An expression vector was designed containing synthetic autoregulated G-Rex or C-

Rex genes and a target promoter. To characterise the functioning of the circuit, a

fluorescent reporter gene, gfp, was put under the control of the target promoter,

allowing several promoter/operator variants to be tested. The constructed vectors

were transformed into G. thermoglucosidasius 11955 Δrex and RNA expression

analysed by qRT-PCR (initial results revealed that GFP fluorescence could not be

used to assess gene expression under oxygen limited conditions).

Design of Rex responsive promoters / operators

Controllable promoters were designed to include a Rex ROP site based on the

EMSA probes previously tested. While most Rex-controlled promoters have single

ROP sites, in Streptomyces double Rex sites have been found upstream of the cyd,

wblE and SCO5207 genes (Brekasis & Paget, 2003; Strain-damerell, n.d.). These

O limited

O limited

O limited

0 30 60 0 30 60 0 30 60 (min)

Figure 5.12 Oxygen limitation results in more abundant Rex protein levels when auto- regulated.

150

double sites overlap one another by 5 bp and therefore occupy opposite sides of the

DNA helix. EMSA experiments using cydA and wblE probes result in double shifts

indicating two Rex dimers bound (Brekasis & Paget, 2003). Importantly, cydA is the

most tightly controlled of the genes in S. coelicolor, perhaps on account of the high

occupancy of the double ROPs site (Brekasis & Paget, 2003). On account of this,

both single and double ROP sites were integrated into promoters. The arrangement

of the G-Rex and C-Rex double ROP sites, with respect to the promoter elements,

were designed based on S. coelicolor cydA and wbIE (Figure 5.13 and 5.14). The

promoters included -35 and -10 regions based on pUP3 with an additional extended

-10 region. Unlike the design of the auto-regulated promoter (section 5.5.1), the

ROP site was positioned downstream of the -10 region similar to the arrangement

in S. coelicolor cydA. Flanking XhoI and XbaI restriction sites were incorporated to

allow cloning into the previously-constructed pG1AK- sfGFP_oriT_G/C-Rex

plasmids (Figure 5.11). The plasmids were conjugated into

G. thermoglucosidasius 11955 Δrex strain.

151

Name Sequence -35 extended -10 ---- Rex site ---- -----

G-rex single ROP

TTGAATGATACCGATGTTTGTGATAGAATTGCTTGTGAAATAATGCACAA

G-rex double ROP

TTGAATGATACCGATGTTTGTGATAGAATTGCTTGTGAATGTGAACACATTCACAA

C-rex single ROP

TTGAATGATACCGATGTTTGTGATAGAATTGCTTGTTAAATTGTTAACAT

C-rex double ROP

TTGAATGATACCGATGTTTGTGATAGAATTGCTTGTTAATGTTAAAACATTAACAA

Figure 5.13 Spacing of double ROP sites including cydA and wbIE found in Streptomyces. Double site sequences designed for G-Rex and C-Rex binding are based on cydA and wbIE. The main interaction points of Rex are indicated in yellow or pink.

Name Sequence

ROP1 ROP2 ROP1 ROP2

Streptomyces cydA Rex ROP site TTGTGAATGTGAACGCGTTCACAA

Streptomyces wblE Rex ROP site TTGTGAATGTGAACGCTTTCACGA

Streptomyces Rex consensus TTGTGAACGCGTTCACAA

TTGTGAACGCGTTCACAA

G-rex double binding site design TTGTGAATGTGAACACATTCACAA

G-Rex consensus TTGTGAAATAATGCACAA

TTGTGAAATAATGCACAA

C-rex double binding site design TTGTTAATGTTAAAACATTAACAA

C-Rex consensus TTGTTAAATAATGAACAA

TTGTTAAATAATGAACAA

Figure 5.14 Design of single and double C/G Rex operator sites. Red font indicates -35 or -10

regions. Blue highlight is extended -10 region. The main interaction points of Rex are indicated in yellow or pink.

152

Study of double and single ROP sites by expression

analysis of gfp reporter

Following the cloning of the various promoter/operator fragments, activity could

be qualitatively assessed using the downstream fluorescence reporter gene gfp. TAS

plates were streaked with G. thermoglucosidasius 11955 Δrex (pG1AK-

sfGFP_oriT_G/C-Rex) strains, and grown at 60 oC overnight. A clear difference in

gfp expression levels could be visually gauged and the fluorescence recorded (Table

5.3)

Name gfp fluorescence strength

G-Rex Single ROPG5/C14 site ++

G-Rex Single ROPT5/A14 site +++

G-Rex Double ROPG5/C14 site +

G-Rex Double ROPT5/A14 site ++

G-Rex constitutive promoter +++

C-Rex Single ROPT5/A14 site +

C-Rex Single ROPG5/C14 site +++

C-Rex Double ROPT5/A14 site -

C-Rex Double ROPG5/C14 site ++

C-Rex constitutive promoter +++

Table 5.3 Visual record of florescence caused by gfp expression. The variety of different gfp promoters are represented, including a strong constitutive promoter. Each promoter was cloned into both C-Rex and G-Rex expression vectors.

As expected the highest fluorescence was observed for the promoters that lacked

ROP sites as well as those that included a non-cognate ROP site (e.g. G-Rex Single

ROPG5/C14 site). However, expression was reduced for both G-Rex and C-Rex

constructs when the promoter contained single cognate ROP sites; ROPG5/C14 and

ROPT5/A14, respectively. In agreement with the EMSA data, this suggests that in vivo

G-Rex binds more strongly to ROPG5/C14 than to ROPT5/A14 and that G-Rex binds

more strongly to ROPT5/A14 than to ROPG5/C14. Furthermore, fluorescence levels

153

were greatly diminished when cognate double ROP sites were used. This suggests

tighter regulation at promoters with double ROP sites, which might be due to

higher degree of Rex occupancy, thereby preventing RNA polymerase binding. A

small decrease in fluorescence was observed between G-Rex Signal ROPT5/A14 and G-

Rex Double ROPT5/A14 sites indicating that G-rex does bind to Crex sites, just less

strongly than to the Grex consensus. Similar results were seen with C-Rex binding

activity and ROPG5/C14.

5.5.4. Response of Rex expression vectors to oxygen limitation

A key aim of this work was to generate a Rex-dependent expression vector that

allowed increased gene expression in response to oxygen limitation. Experiments

were therefore performed where G. thermoglucosidasius 11955 Δrex (pG1AK-

sfGFP_oriT_G/C-Rex) strains were grown in oxygen abundant, or limited

conditions. A potential problem was that the gfp reporter requires molecular

oxygen for the maturation of the fluorophore. Nonetheless gfp has been used for

similar experiments in the past by allowing oxygen-dependent maturation following

the inhibition of further protein synthesis (Roggiani & Goulian, 2015). In our

hands, cultures grown under oxygen limitation had very low levels of fluorescence

and attempts to allow chromophore maturation did not work. This might be

because this gfp is a superfolder with increased resistance to denaturation, but

becomes trapped in a non-fluorescent folding intermediate state (Dammeyer &

Tinnefeld, 2012; Fisher & DeLisa, 2008). Therefore an alternative approach of

using qRT-PCR to measure gfp transcript abundance was used; a benefit of this

approach is that the expression of other known Rex targets within the genome

could also be assessed.

To confirm previous visual qualitative assessment of gfp expression, strains were

grown aerobically at 250 rpm in shaking flasks to OD600 1.0-1.5, and a 1.5 ml

sample was taken of each culture at time point 0 (oxygen abundant time point) and

processed to isolate RNA and qRT-PCR performed. Results in Figures 5.15 and

5.16 indicated that gfp transcript abundance was reduced for nearly all G-Rex and

C-Rex constructs, compared to the constitutive promoter (apart from C-Rex double

ROPG5/C14 site, this was negligibly higher). A substantial decrease in gfp

transcript abundance was measured between C-Rex single/double ROPT5/A14 site

154

and C-Rex single/double ROPG5/C1 indicating that C-Rex binds more strongly to

ROPT5/A14 than ROPG5/C1 sites (Figure 5.15). This was not reciprocal in G-Rex

constructs, and no significant difference in gfp transcript abundance was measured

between ROPT5/A14 and ROPG5/C1 (Figure 5.16). gfp transcript abundance decreased

in cognate double ROP T5/A14 sites, but not double ROPG5/C1 sites (compared to

single ROPs) in both G-Rex and C-Rex constructs. The results suggest that the

double ROP T5/A14 site is more favourable to rex regulation, potentially permitting

two Rex dimers to bind simultaneously, producing tighter control over gene

expression. The viable qualitatively florescence results may differ slightly from

qRT-PCR data potentially due to the fast turnover of mRNA, whereas gfp protein

builds up over a far longer period of time.

To analyse if the activity of Rex is responsive to changing oxygen abundance in

vivo, strains were grown as stated before but after time point 0 the remaining

culture was used to fill 1.5 ml microfuge tubes to the top, so no air remained. Tubes

were then incubated without shaking to produce oxygen limited growth conditions

and harvested every 5 min for a total of 25 min, then processed to isolate RNA and

qRT-PCR performed. gfp transcript abundance increased as time under oxygen

limited conditions increased (Figure 5.17). This indicates both G-Rex and C-Rex

sense a change in redox caused by limited oxygen and associate with NADH,

causing disassociation with ROP sites. gfp expression levels were also overall higher

in G-Rex constructs than C-Rex, suggesting that C-Rex binds more tightly to the

ROP T5/A14 site than G-Rex binds the ROPG5/C1 site.

155

Figure 5.15 C-Rex binds to ROPT5/A14 site more strongly than to ROPG5/C14 site. qRT- PCR of gfp target with C-Rex, both double and single G/C rex ROP. As a control gfp was put

under the control of a constitutive promoter. All samples are of time point 0, oxygen abundant growth conditions. Graph is a logarithmic scale. Data was normalised against pgi transcript

abundance. Error bars are of standard deviation and data represents two biological repeats.

Figure 5.16 G-Rex binds to ROPG5/C14 site and ROPT5/A14 site equaly. qRT-PCR of gfp target with G-Rex, both double and single G/C rex ROP. As a control gfp was put under the

control of a constitutive promoter. All samples are of time point 0, oxygen abundant growth conditions. Graph is a logarithmic scale. Data was normalised against pgi transcript abundance. Error bars are of standard deviation and data represents two biological repeats.

156

5.5.5. Analysis of chromosomal Rex targets in response to oxygen

limitation

These experiments were set up to establish if the G. thermoglucosidasius 11955 Δrex

(pG1AK-sfGFP_oriT_G/C-Rex) strains are capable of regulating known rex targets

in vivo. Strains containing G-Rex Single ROPG5/C14 site and C-Rex Single

ROPT5/A14 site were grown as stated before, in oxygen limited conditions. The

transcript abundance of adhE and ldh were measured over a time course and the

results presented in Figure 5.18 and 5.19. Results show that adhE becomes more

highly expressed with increasing oxygen limitation in both strains, as expected

(Figure 5.18). However, adhE becomes far overexpressed in the C-Rex strain, rising

to a transcript abundance of over 1200, compared to G-Rex which reached a height

of just over 200. This indicates that G-Rex binds more strongly to the genomic ROP

(min)

G-Rex Single ROPG5/C14 site

C-Rex Single ROPT5/A14 site

Figure 5.17 Oxygen limitation affects C-Rex and G-Rex binding. qRT-PCR of gfp with two

strains, G. thermoglucosidasius 11955 Δrex (p11AK2_oriT_G/C-Rex) strains with over time points 0-25 min. At time point 0 samples were taken during oxygen abundant growth conditions. Samples 5-25 min were increasingly oxygen limited. Graph is a logarithmic scale. Data was normalised against pgi transcript abundance. Error bars are of standard deviation representative of two biological

repeats.

157

site than C-Rex, and is therefore able to inhibit expression to a greater degree. Ldh

transcript abundance is more highly expressed in the C-Rex strain than G-Rex

(Figure 5.19) indicating again that G-Rex is more able to bind G. thermoglucosidasius

genomic ROP sites than C-Rex. These results demonstrate the orthogonal nature of

C-Rex in G. thermoglucosidasius bacterium.

158

adhE

(min)



Figure 5.18 qRT-PCR of adh target with two strains, G/C-Rex_single ROP over time points 0-25 min. At time point 0 samples were taken during oxygen abundant growth conditions.

Samples 5-25 min were increasingly oxygen limited. Data was normalised against pgi transcript abundance. Error bars are of standard deviation and data represents two biological repeats.

ldh

(min)



Figure 5.19 qRT-PCR of ldh target with two strains, G/C-Rex_single ROP over time points 0-

25 min. At time point 0 samples were taken during oxygen abundant growth conditions. Samples 5-25 min were increasingly oxygen limited. Data was normalised against pgi transcript abundance. Error bars are of standard deviation and data represents two biological repeats.

159

5.5.6. Complementation of the rex mutant

Experiments were conducted to establish if the G. thermoglucosidasius 11955 Δrex

(pG1AK-sfGFP_oriT_G/C-Rex) strains are capable of Δrex complementation.

Cultures were grown of the 11955 WT, 11955 Δrex, 11955 Δrex (pG1AK-

sfGFP_oriT_G-Rex Single ROPG5/C14) and 11955 Δrex (pG1AK-sfGFP_oriT_C-Rex

Single ROPT5/A14) in both oxygen abundant (T=0) and limited (T=30) conditions,

processed to isolate RNA and qRT-PCR performed on known Rex targets. In both

Figures 5.20 and 5.21 the expression profile of WT matches that of the G-Rex

complement. Regulation of adhE and ldh was restored by the pG1AK-

sfGFP_oriT_G-Rex Single ROPG5/C14 plasmid almost completely. A slight

elevation of transcript abundance was present and this may be due to the difference

in rex expression, as the WT was constituently expressed compared to the

complement G-Rex being auto-regulated. These results indicate that the 11955 Δrex

(pG1AK-sfGFP_oriT_G-Rex Single ROPG5/C14) Complement strain was capable of

controlling rex target gene expression.

The transcript abundance of adhE and ldh Rex targets was extremely elevated in the

11955 Δrex (pG1AK-sfGFP_oriT_C-Rex Single ROPT5/A14) strain compared to WT,

and matches closely to the Δrex strain (Figure 5.20 and 5.21). These results indicate

that C-Rex was not able to bind genomic ROP sites of adhE and ldh targets to

regulate the expression of these genes. This demonstrates that C-Rex is orthogonal

gene regulator in G. thermoglucosidasius.

160

adhE

(min)

Figure 5.20 G-Rex, but not C-Rex complements a Δrex. qRT_PCR of adhE target using RNA RNA

isolated from four strains Δrex, wild type 11955 and G. thermoglucosidasius 11955 Δrex (p11AK2_oriT_G/C-Rex) strains. Time point 0 was taken during arobic respiration and time pint 30 min was taken during oxygen limited growth conditions. Data was normalised against pgi transcript

abundance. Error bars are of standard deviation and data represents two biological repeats

ldh

(min)

Figure 5.21 G-Rex, but not C-Rex complements a Δrex. qRT_PCR of ldh target using RNA

isolated from four strains Δrex, wild type 11955 and G. thermoglucosidasius 11955 Δrex (p11AK2_oriT_G/C-Rex) strains. Time point 0 was taken during arobic respiration and time pint 30 min was taken during oxygen limited growth conditions. Data was normalised against pgi transcript abundance. Error bars are of standard deviation and data represents two biological repeats

adhE

ldh

161

5.6. Discussion and conclusions

Prior to this study Clostridia spp. had been predicted to have an alternative ROP

sequence (TTGTTAANNNNTTAACAA) compared to that found in most other

groups of bacteria (TTGTGAANNNNTTCACAA) (D. A. Ravcheev et al., 2012).

This raised the possibility that a Rex orthologue would be able to discriminate

between such ROP sites that differ only by two positions: ROPG5/C14 and ROPT5/A14,

which might allow the development of orthologous expression vectors. However,

since Clostridium thermocellum contains two paralogues, an initial aim was to

determine whether the Cthe_0422 and Cthe_1798 paralogues were each active in

DNA binding and redox sensing. Surprisingly, Cthe_0422 Rex protein did not bind

to any of the ROP sites tested. This was particularly unusual as Cthe_0422 was

predicted to be co-transcribed with bi-functional acetaldehyde-CoA/alcohol

dehydrogenase, a gene that is part of the core Rex regulon across many taxonomic

fnr

Figure 5.22 G-Rex, but not C-Rex complements a Δrex. qRT_PCR of fnr target using RNA

isolated from four strains Δrex, wild type 11955 and G. thermoglucosidasius 11955 Δrex

(p11AK2_oriT_G/C-Rex) strains. Time point 0 was taken during arobic respiration and time pint 30 min was taken during oxygen limited growth conditions. Data was normalised against

pgi Error bars are of standard deviation and data represents two biological repeats.

162

groups. A similarly non-functioning Rex protein, TM1427. was previously found in

Thermotogales maritima, which did not bind to any predicted ROP, and a

functioning paralog TM0169 was also found (D. A. Ravcheev et al., 2012). The

TM1427, TM0169, Cthe_0422 and Cthe_1798 protein sequences were aligned and

it seems likely that TM1427 and Cthe_0422 are orthologues, more closely related to

one another than to the paralogs they share a genome with.

Unlike Cthe_0422, Cthe_1798 did bind to all the EMSA probes, although there was

a clear preference for ROPT5/A14 compared to ROPG5/C14. Furthermore it was found

that G-Rex demonstrated more binding activity to ROPG5/C14 than to ROPT5/A14.

These results agree with the predictions and suggested that the ROP of Clostridia,

combined with the Rex paralog Cthe_1798 could be used as an alternative and

possibly orthogonal redox responsive regulatory system in G. thermoglucosidans.

Such a system would offer relatively similar and related regulation, yet

independent. To test this in vivo C/G-Rex-based genetic circuit vectors were

constructed and transformed into a Δrex background strain. A gfp reporter gene was

put under the control of either signal/double ROPT5/A14 or ROPG5/C14 sites. Results

indicated that C-Rex acted as an orthologue, inhibiting expression of gfp at

ROPT5/A14 far more than at ROPG5/C14 sites. C-Rex also appeared to be capable of

distinguishing single from double sites as gfp was more tightly inhibited at double

ROP sites. Results indicated that G-Rex was not as able at distinguishing between

single and double sites as gfp was expressed at near equal amounts. G-Rex did not

demonstrate differential binding between ROPT5/A14 and ROPG5/C14 sites as gfp

expression did not differ significantly. These results are promising as alterations

such as the replacement of ROPG5/C14 for ROPT5/A14 sites, and the introduction of

double sites may have no effect on G-Rex regulation, but C-Rex could differentiate

between such sites readily. C-Rex and G-Rex were demonstrated to both be reactive

to changing oxygen abundance, as gfp expression increased with longer duration of

oxygen starvation, and are therefore likely to be able to sense cell redox.

Genomic Rex targets adhE and ldh were also analysed and it was found that

expression of G-Rex by pG1AK-sfGFP_oriT_G-Rex plasmids was capable of

complementing the G. thermoglucosidasius 11955 rex mutant strain. Figure 5.6.1

illustrates how regulators Rex, Fnr and ResDE are thought to interact after

conducting this research. C-Rex expression in G. thermoglucosidasius 11955 rex

mutant strain was not, however, capable of

163

complementing, and the expression of Rex targets adhE and ldh mimicked that of

the mutant, thereby demonstrating again that C-Rex is a likely orthologue.

The genetic circuit vectors could be diversified further to experiment with the

expression of Rex. During this project we demonstrated that G. thermoglucosidasius

wild type Rex expression is likely constitutive. To optimise previous research and

generate orthogonal Rex systems, a major focus of research should be of the control

over the expression of the orthogonal Rex itself, especially once integrated into the

genome. Inducible promoter systems offer robust control, however the cost of

adding large quantities to a bioreactor quickly becomes expensive and is not cost

effective. Autoregulation proved to be a successful mechanism to prevent the over

expression of Rex when on multicopy plasmids and utilising autoregulation is likely

to be the most robust, dynamic and cost effective mechanism of orthogonal C-Rex

control. If the strength of the promoter and the tightness of Rex regulation can be

balanced, as it was in the design of pG1AK-sfGFP_oriT_G/C-Rex plasmids, then

this may offer the best solution. Perhaps research can be accelerated by studying the

similarities between previously demonstrated Rex autoregulation (Bitoun et al.,

2012; Brekasis & Paget, 2003) and also those predicted to be (D. A. Ravcheev et al.,

2012).

The use of orthogonal systems to divide gene expression into independent modules

can aid in achieving predictable outcomes and optimisation of product yield. Cells

and biological systems are interconnected and any change in one pathway is likely

to have unpredictable functional or temporal effects on another (Kavscek, Strazar,

Curk, Natter, & Petrovc, 2015). For this reason orthogonality is not absolute, but

instead a sliding scale. Many varieties of orthogonal systems have been developed,

some centring on generating promoter libraries but others are far more ambitious,

and some aim expand the genetic code (Blount, Weenink, Vasylechko, & Ellis,

2012; Gilman & Love, 2016; Maranhao & Ellington, 2016). The development of an

orthogonal system makes it possible to engineer the cell metabolism. To gain more

control over the expression of key enzymes, such as those specifically involving

fermentation is a powerful way to divert metabolic flux towards the product of

interest, ethanol. Here we have demonstrated the orthogonal potential of C-Rex

and expanded the repertoire of ROP sites that could be used in such a system.

Further work could explore how C-Rex and G-Rex might compete for ROPT5/A14

164

sites. This would help to develop a more sophisticated understanding of how C-Rex

could be used in an orthogonal system. It seems likely that C-Rex would ‘override’

G-Rex for ROPT5/A14 sites but not ROPG5/C14 sites and specific genes could be

targeted for tighter regulation. An understanding of the exact redox sensing

properties of both G-Rex and C-Rex would enable prediction as to what cellular

and environmental conditions would result in association, or dissociation with

ROP sites, giving greater insight into how the orthogonal system could be

implemented in large bioreactors to trigger gene expression. The orthogonal C-Rex

system analysed during this project has wider applications beyond G.

thermoglucosidasius, and could be applied in any of the many bacteria conforming to

the typical ROPG5/C14 consensus sequence (D. A. Ravcheev et al., 2012).

Figure 5.6.1 Gene regulatory network. Rex binding to ROP sites is induced by NADH, when bound it

inhibits expression of ldh, adh and fnr. Fnr competes with Rex for ROP sites and positively regulates

expression of adh, and auto regulates itself. ResDE is active under oxygen limited conditions and induces

expression of fnr.

165

6. Chapter VI: Discussion

Overview

To produce biofuel in the most efficient way possible it is important to fully

understand the regulators controlling fermentation and anaerobic metabolism, and

the factors triggering their activation. Rex is a key regulator of fermentation genes,

able to sense redox balance within the cell. By understanding and manipulating Rex

it is anticipated that fermentation can be more precisely controlled and the

expression of genes optimised for industrial bioreactors. Rex evolved as a

mechanism to cope with redox stress caused, for example, by oxygen deprivation,

often faced by soil-dwelling bacteria. Rex is found in many Gram-positive bacteria

and predictions of regulons based on the conserved ROP operator have revealed

that certain “core” genes are particularly prevalent. One example is the bifunctional

acetaldehyde-CoA/alcohol dehydrogenase (ADH) (D. a Ravcheev et al., 2012).

ADH converts acetyl-coenzyme A to ethanol via the intermediate acetaldehyde,

and oxidises two NADH molecules to NAD+ in the process, thereby helping to

maintain the redox balance. It is vital for the cell to recycle NADH in the absence

of oxygen, as the glycolytic pathway depends upon NAD+ to generate ATP. Only

during a build-up of NADH (usually occurring during oxygen limited growth

conditions) would a less efficient form of energy production, fermentation, be

necessary. Rex is a direct sensor of NADH/NAD+ ratio (Brekasis & Paget, 2003)

and is therefore uniquely placed to regulate fermentation and anaerobic respiration.

Much work has been carried out in industrially important Clostridium to this effect

(Wietzke & Bahl, 2012b; L. Zhang et al., 2014). In C. acetobutylicum Rex was shown

to control adhE and ldh genes as well as the bca operon encoding genes important

for the C4 metabolic pathway (Wietzke & Bahl, 2012). Predicted Rex target genes in

a range of Clostridium species include adhA, adhE, crt-bcd-etfBA-hbd and ptb-buk

operon encoding enzymes important for alcohol synthesis (L. Zhang et al., 2014).

Other predicted Rex controlled genes included those involved in sulphite and

nitrate reduction, hydrogen production, fermentation, the TCA cycle, and NAD

biosynthetic genes, and some targets were confirmed by EMSA (L. Zhang et al.,

2014). This project focused on G. thermoglucosidans, which also has great industrial

potential, but has an underdeveloped genetic manipulation toolkit. The aim was to

166

investigate the G. thermoglucosidans Rex regulon, explore the use of a clostridial Rex

as an orthogonal regulator for use in Geobacillus, and to construct new tools for

genetic manipulation.

6.1. General discussion and future directions

The Rex regulon of G. thermoglucosidans was investigated by Chip-seq and qRT-

PCR and was found to include genes that encode enzymes involved in

fermentation, anaerobic respiration, and carbon metabolism, as well as several

genes with no known function. ChIP-seq also identified a number of binding sites

that are internal to genes; while Rex is unlikely to regulate these genes it is

conceivable that Rex regulates sRNAs that are encoded with the genes.

Interestingly, Rex itself was not auto-regulated, unlike the case in Streptomyces

coelicolor (Brekasis & Paget, 2003), but rather appears to be unaffected by oxygen

limitation.

Much work has been focused on identifying members of the rex regulon, however

these have focused on the use of bioinformatics and a consensus sequence. The

study conducted here is the first to apply ChIP-sequencing to identify Rex binding

locations at a gnomically global level. Rex was previously shown to be important

for the control of fermentation genes in a variety of Gram positive bacteria

including Staphlococcus aureus, strains of Clostridia and Bacillus (Pagels et al., 2010;

D. a Ravcheev et al., 2012; Wang, Bauer, Rogstam, Linse, Logan, & von

Wachenfeldt, 2008; L. Zhang et al., 2014). The predicted conserved Rex regulon

across 11 taxonomic groups of bacteria included adhE, adhA, adhB, ldh, pflAB, dhaT,

butA and bcd-hbd (D. a Ravcheev et al., 2012). Some of these fermentative Rex

targets such as ldh and adhE have been confirmed experimentally in G.

thermoglucosidans during this study. The G. thermoglucosidans Δrex strain constructed

during this study was predicted to produce high amounts of ethanol, however this

was not what our results showed. A Δrex strain of C. acetobutylicum had been

demonstrated to increase yields of butanol and ethanol, and to start production

much earlier than in a WT strain (Wietzke & Bahl, 2012). Deregulation in our G.

thermoglucosidans Δrex strain did lead to an increases in yield and earlier production

of one fermentation product, lactate. This outcome could possibly be improved by

167

the generation of a double mutant (Δrex, Δldh) to redirect flux towards the ADH

pathway. A G. thermoglucosidans Δldh has been previously analysed for the

production of metabolites and it was found that although ethanol yields increased,

the remaining pathways were not able to utilise the increase in carbon flux and

pyruvate built up (Cripps et al., 2009). To deal with this the PDH gene was

upregulated which resulted in almost a doubling in ethanol yield (Cripps et al.,

2009). Our results indicate that Rex deregulation results in increased carbon flux

towards LDH, however if LDH was deleted this may have the effect of redirecting

it again towards ADH. Since adhE is overexpressed in a Δrex, this strain might be

able to cope with an in increased carbon flux well, especially if combined with an

upregulated PDH.

Genes involved in carbon metabolism and glycolysis such as gldA, gap, pgk, tpi and

fba have previously been described as Rex targets (D. a Ravcheev et al., 2012). We

discovered that Rex binds upstream of an operon encoding 6-phosphofructo kinase

(pfkA) and pyruvate kinase (pyK), key enzymes in the glycolytic pathway. PFK is

particularly important as it controls the rate limiting step in glycolysis and therefore

plays a central role in carbon flux. In E. coli PFK is controlled in response to levels

of the product fructose 6-phosphate, by the central carbon pathways regulator Cra,

and in Bacillus CcpA (Shimada, Yamamoto, & Ishihama, 2011; Tobisch et al.,

1999). In addition the adhE gene in E. coli was found to also be controlled by Cra

(Mikulskis, Aristarkhov, & Lin, 1997). PYK is controlled by Cra and CcpA, and in

Bacillus PYK has been shown to be involved in acetate production and a Δpyk

demonstrated significant reduced in organic acids (Fry et al., 2000). The overall role

of CcpA includes the activation of glycolysis in B. subtitlis and response to changing

conditions by directing carbon overflow metabolism. Rex may control pfkA and pyK

expression in conjunction with CcpA to help to direct flux and address redox

imbalance.

Redox imbalance is also adjusted by the expression of ndh, shown to be a Rex target

in G. thermoglucosidans during this study. ndh is a NADH dehydrogenase and in B.

subtilis is also regulated by Rex forming a regulatory loop that avoids large

fluctuations in the ratio of NADH/NAD+ within the cell (Gyan et al., 2006).

168

This study has also demonstrated that Rex binds upstream of nitrite reductase

(narK) and a nitrite transporter (nirC) in G. thermoglucosidans; both genes were

previously identified as putative Rex targets in a variety of bacteria (D. a Ravcheev

et al., 2012). A number of other anaerobic, nitrite reduction genes such as nark and

nirK were also identified as potentially controlled by Rex during ChIP-seq analysis.

NirK is a copper containing nitrite reductase found in several denitrifying bacteria,

and NarK a nitrite extrusion protein used in protecting cells from stress of nitrite

within the cytoplasm, also regulated by Fnr in B. subtitlis (Bueno et al., 2012;

Rodionov, Dubchak, Arkin, Alm, & Gelfand, 2005). The data suggests that Rex

might control nitrite dissimilation, which could be a way of addressing redox

balance. An alternative possibility, given the close location of fnr to nark and nirK,

and the fact that these enzymes are controlled by Fnr in Bacillus, is that the binding

seen is spurious, possibly caused by the similarity between binding sites. B. subtilis

Fnr consensus binding site is very similar (TGTGA-n6-TCACT) to the Rex

consensus (TTGTG-n8-CACAA) (Reents et al., 2006; Wang, Bauer, Rogstam,

Linse, Logan, & von Wachenfeldt, 2008). Is it also possible that Rex and Fnr have

overlapping specificity and control the some of the same genes. It could be that Rex

modulates expression under anaerobic conditions, depending on redox state of the

cell, and Fnr induces gene expression in an oxygen limited environment. Data to

support this demonstrated that neither adhE nor ldhA are constitutively expressed in

a rex mutant, indicating other regulators induce expression of these genes,

regulators such as Fnr, or ResDE. A Rex binds upstream of other genes important

for nitrate reductase including fnr/modCD operon and a mobA gene, discovered

during this study. Mob converts molybdopterin to molybdopterin guanine

dinucleotide which is require for nitrate reductase and rex shares an operon with a

gene involved in molybdopetrin production (moaC). Data indicates that Rex is a

regulator of not only genes involved in anaerobic respiration and nitrate reduction,

but also an important inducer of anaerobic respiration, the regulator Fnr. ChIP-seq

revealed that fnr does appear to be directly controlled and that Rex deregulation did

have a significant impact on fnr expression, but this was confined to oxygen limited

growth conditions. Interestingly the expression pattern in a Δrex strain mimicked

that of adhE, however other Rex targets such as ldh and ndh did not match this

expression profile. The differing expression profiles of genes such as ldh and adhE in

169

a deregulated Δrex strain indicated that other regulators may also act upon Rex

targets. Perhaps future research should focus on ascertaining if the regulators Rex

and Fnr compete (or bind together at double ROP sites) to fine-tune gene

expression in response to environmental oxygen, and cellular redox conditions.

Experiments should focus on whether Rex may act to dampen Fnr gene expression

induction, especially in relation to the adhE gene.

An alignment of G-Rex binding sites identified using Chip-seq data allowed the

generation of a G-Rex ROP consensus sequence with the most conserved residues

including the GTG-n8-CAC motif. The Rex binding site is extremely well

conserved among Gram-positive bacteria and extends to the remaining residues in

the 16 bp site. However, the prediction that clostridial Rex proteins bind with an

altered specificity (GTT n8 AAC) was exploited here to develop a Clostridium

thermocellum Rex (C-Rex)-based expression system that might be orthogonal in

Geobacillus. The data presented indicated that C-Rex was expressed but did not

restore regulation in a Δrex strain, which suggested that it might not be binding to

host Rex binding sites and therefore might be orthogonal.

An orthogonal regulator opens up the possibility for control over fermentation

regulation. ROP sites of specific genes could be modified and optimised for C-Rex

binding, so that altering C-Rex levels would give control over these genes or any

new genes which might be integrated. C-Rex potentially has a different affinity for

NADH and genes could be switched on or off at slightly altered redox potentials. In

addition to this a library of ROP sites, including single and double sites, could be

produced that range in strength. Approaches in the past have generated promoter

libraries to construct orthogonal genetic circuits (Stanton et al., 2014). This would

allow a form of temporal control over the expression of genes in response to redox.

We have shown that the design of an auto regulating Rex expression system

functions efficiently and can dynamically respond to changing conditions. Other

systems have utilised dynamic regulatory networks to optimise the production of

fatty acid biosynthesis (Xu, Li, Zhang, Stephanopoulos, & Koffas, 2014). Inducible

promoters are not cost effective in an industrial setting, but auto regulation does

provide a real alternative.

170

C-Rex was proven to exhibit a greater affinity for an alternative ROP site than to

the predicted G. thermoglucosidans consensus sequence. G-Rex did however bind to

both its own, and to C-Rex ROP sites. These experiments were conducted in vitro

and vivo and C-Rex was demonstrated to respond not only to NADH levels but

also NAD+ (which enhanced binding to ROP sites). Experiments also included

gene regulatory networks that incorporated auto-regulated G-rex and C-Rex, which

was successful in dynamically responding to changing environmental conditions

(oxygen abundance). The orthogonal system could be used in the future to optimise

the expression of particularly significant genes, by replacing the original ROP site

for that of C. thermocellum, thereby creating an independent redox responsive

regulatory system separate form G-Rex. In addition to the use of C-Rex as an

expression control aid, double ROP sites were demonstrated to lead to tighter

regulation than single sites. In combination, orthogonal C-Rex, autoregulation and

double ROP sites could expand the repertoire of the rex regulatory system and lead

to precise modulation of fermentation.

The cloning and transfer of large (~80kb) biosynthetic clusters remains challenging,

even in model organisms (Nah, Woo, Choi, & Kim, 2015). Optimisation of DNA

transfer will be important for the genetic engineering of G. thermoglucosidans

genome. Before work began to characterise the Rex regulon or explore C-rex as an

orthogonal regulator, genetic manipulation tools were constructed. Much work was

done to generate an integrative vector system, including the construction of an

integrative vector, however this was unsuccessful. The placement of an attachment

site within the G. thermoglucosidans genome proved difficult due to the combining of

this aim with another, for the production of a counter selection system. Future

work would build upon the pyrE counter selection system to construct a ΔpyrE

ΔpryR double mutant strain for counter selection to be successful. The counter

selection system could then be used to insert an attachment site into the G.

thermoglucosidans genome. An alternative integrative vector system could utilise the

φBT1 phage attachment sites instead of φC31 (Gregory, Till, & Smith, 2003).

Functional selection markers are also limiting, as currently only two thermostable

antibiotic markers are available for use in G. thermoglucosidans (Reeve et al., 2016).

The development of additional antibiotic marks needs both a thermostable

antibiotic with a long half-life, and also an antibiotic resistance gene active under

171

high growth temperatures of 60oC. This is difficult to achieve: Kanamycin is a

commonly used antibiotic as it has good thermostability but standard mesophilic

kanamycin resistance genes encode enzymes which are not thermostable. Therefore

several thermostable resistant derivatives were developed by forced evolution (H.

Liao, McKenzie, & Hageman, 1986). Gentamicin, a member of the

aminoglycosides class of antibiotics, is effective against Gram positive bacteria and

relatively thermostable and often used as a bacterial agent in cell culture (Fischer,

1975). A gentamicin resistance has been developed for use in cloning and may not

need to undergo forced evolution to work in G. thermoglucosidans (Bian, Fenno, &

Li, 2012; Poggi, Oliveira de Giuseppe, & Picardeau, 2010).

Prior to this project DNA transformation was inefficient and required specialist

equipment, hampering progress. A conjugative method was developed (including

strains and plasmids) which facilitated our research. Other research had also

produced a working conjugative system for use with Geobacillus kaustophilus, which

was then expended and tested in other Bacillus and Geobacillus species (Suzuki &

Yoshida, 2012; Tominaga, Ohshiro, & Suzuki, 2016). Non-replicating conjugative

vectors were constructed and used for gene disruption (to produce a Δrex strain) and

gene replacement (3xFLAG tagged Rex Strain). Optimisation of the conjugation

protocol led to the use of magnesium supplementation and dcm- donor E. coli strain.

Results indicated that methylation did play a role in transformation frequency, but

whether or not dcm is the only system to be important is yet to be discovered.

Throughout experimentation with G. thermoglucosidans it become clear that it is

sensitive to cold shock at room temperature and that minimising this leads to

enhanced transformation. A modular origin of transfer component was also

constructed for use with the modular plasmid toolkit set (Reeve et al., 2016), which

was utilised during the analysis of the orthogonal C-Rex protein.

This thesis has made a novel contribution to the understanding of the Rex regulon,

and significantly expanded the genetic engineering tools that enable strain

development. The work presented has demonstrated the potential of Rex to be

used in an othoganol system for second generation biofuel production. Many more

questions have arisen during the course of our research that if pursued could

potentially aid the progress of G. thermoglucosidans, or other gram negative bacteria,

to be successful biocommodities producers.

172

7. Bibliography

Abdel-Rahman, M. A., Tashiro, Y., & Sonomoto, K. (2010). Lactic acid

production from lignocellulose-derived sugars using lactic acid bacteria:

overview and limits. Journal of Biotechnology, 156(4), 286–301.

doi:10.1016/j.jbiotec.2011.06.017

Alexeeva, S., Hellingwerf, K. J., & Teixeira de Mattos, M. J. (2003). Requirement

of ArcA for Redox Regulation in Escherichia coli under Microaerobic but Not

Anaerobic or Aerobic Conditions. Journal of Bacteriology, 185(1), 204–209.

doi:10.1128/JB.185.1.204-209.2003

Alting-Mees, M. A., & Short, J. M. (1989). pBluescript II: gene mapping vectors.

Nucleic Acids Research, 17(22), 9494. Retrieved from

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=335171&tool=p

mcentrez&rendertype=abstract

Alvira, P., Tomás-Pejó, E., Ballesteros, M., & Negro, M. J. (2010). Pretreatment

technologies for an efficient bioethanol production process based on enzymatic

hydrolysis: A review. Bioresource Technology, 101(13), 4851–61.

doi:10.1016/j.biortech.2009.11.093

Armaroli, N., & Balzani, V. (2011). The legacy of fossil fuels. Chemistry, an Asian

Journal, 6(3), 768–84. doi:10.1002/asia.201000797

Baruah, A., Lindsey, B., Zhu, Y., Michiko, M., Baruah, A., Lindsey, B., …

Nakano, M. M. (2004). Mutational Analysis of the Signal-Sensing Domain of

ResE Histidine Kinase from Bacillus subtilis Mutational Analysis of the Signal-

Sensing Domain of ResE Histidine Kinase from Bacillus subtilis.

doi:10.1128/JB.186.6.1694

Bateman, J. R., & Wu, C. (2008). A simple polymerase chain reaction-based

method for the construction of recombinase-mediated cassette exchange donor

vectors. Genetics, 180(3), 1763–6. doi:10.1534/genetics.108.094508

Bechhofer, D. H., & Wang, W. (1998). Decay of ermC mRNA in a polynucleotide

phosphorylase mutant of Bacillus subtilis. Journal of Bacteriology, 180(22), 5968–

77. Retrieved from



Bekker, M., Alexeeva, S., Laan, W., Sawers, G., Teixeira de Mattos, J., &

Hellingwerf, K. (2010). The ArcBA two-component system of Escherichia coli

is regulated by the redox state of both the ubiquinone and the menaquinone

pool. Journal of Bacteriology, 192(3), 746–54. doi:10.1128/JB.01156-09

Berks, B. C., Sargent, F., & Palmer, T. (2000). The Tat protein export pathway.

Molecular Microbiology, 35(2), 260–274. doi:10.1046/j.1365-2958.2000.01719.x



173

Bhalla, A., Bansal, N., Kumar, S., Bischoff, K. M., & Sani, R. K. (2013). Improved

lignocellulose conversion to biofuels with thermophilic bacteria and

thermostable enzymes. Bioresource Technology, 128, 751–9.

doi:10.1016/j.biortech.2012.10.145

Bian, J., Fenno, J. C., & Li, C. (2012). Development of a modified gentamicin

resistance cassette for genetic manipulation of the oral spirochete Treponema

denticola. Applied and Environmental Microbiology, 78(6), 2059–62.

doi:10.1128/AEM.07461-11

Bidnenko, V., Shi, L., Kobir, A., Ventroux, M., Pigeonneau, N., Henry, C., …

Mijakovic, I. (2013). Bacillus subtilis serine/threonine protein kinase YabT is

involved in spore development via phosphorylation of a bacterial recombinase.

Molecular Microbiology, 88(5), 921–35. doi:10.1111/mmi.12233

Bierman, M., Logan, R., O’Brien, K., Seno, E. T., Rao, R. N., & Schoner, B. E.

(1992). Plasmid cloning vectors for the conjugal transfer of DNA from

Escherichia coli to Streptomyces spp. Gene, 116(1), 43–9. Retrieved from

http://www.ncbi.nlm.nih.gov/pubmed/1628843

Bitoun, J. P., Liao, S., Yao, X., Xie, G. G., & Wen, Z. T. (2012). The Redox-

Sensing Regulator Rex Modulates Central Carbon Metabolism, Stress

Tolerance Response and Biofilm Formation by Streptococcus mutans. PLoS

ONE, 7(9). doi:10.1371/journal.pone.0044766

Bittner, F., & Mendel, R. R. (2010). Cell biology of molybdenum. Plant Cell

Monographs, 17, 119–143. doi:10.1007/978-3-642-10613-2_6

Blount, B. a, Weenink, T., Vasylechko, S., & Ellis, T. (2012). Rational

diversification of a promoter providing fine-tuned expression and orthogonal

regulation for synthetic biology. PloS One, 7(3), e33279.

doi:10.1371/journal.pone.0033279

Boeke, J. D., La Croute, F., & Fink, G. R. (1984). A positive selection for mutants

lacking orotidine-5′-phosphate decarboxylase activity in yeast: 5-fluoro-orotic

acid resistance. Molecular and General Genetics MGG, 197(2), 345–346.

doi:10.1007/BF00330984

Bogner, J., Pipatti, R., Hashimoto, S., Diaz, C., Mareckova, K., Diaz, L., …

Gregory, R. (2008). Mitigation of global greenhouse gas emissions from waste:

conclusions and strategies from the Intergovernmental Panel on Climate

Change (IPCC) Fourth Assessment Report. Working Group III (Mitigation).

Waste Management & Research, 26(1), 11–32. doi:10.1177/0734242X07088433

Bolhuis, A., Sorokin, A., Azevedo, V., Ehrlich, S. D., Braun, P. G., deJong, A., …

Maarten van Dijl, J. (1996). Bacillus subtilis can modulate its capacity and

specificity for protein secretion through temporally controlled expression of the


174

sipS gene for signal peptidase I. Molecular Microbiology, 22(4), 605–618.

doi:10.1046/j.1365-2958.1996.d01-4676.x

Borisov, V. B., Gennis, R. B., Hemp, J., & Verkhovsky, M. I. (2011). The

cytochrome bd respiratory oxygen reductases. Biochimica et Biophysica Acta -

Bioenergetics, 1807(11), 1398–1413. doi:10.1016/j.bbabio.2011.06.016

Borisov, V. B., & Verkhovsky, M. I. (2009). Oxygen as Acceptor. EcoSal Plus, 3(2).

doi:10.1128/ecosalplus.3.2.7

Brekasis, D., & Paget, M. S. B. (2003). A novel sensor of NADH / NAD + redox

poise in Streptomyces coelicolor A3 ( 2 ). EMBO Journal, 22(18).

Brumm, P. J., Land, M. L., & Mead, D. A. (2015). Complete genome sequence of

Geobacillus thermoglucosidasius C56-YS93, a novel biomass degrader isolated

from obsidian hot spring in Yellowstone National Park. Standards in Genomic

Sciences, 10(1), 73. doi:10.1186/s40793-015-0031-z

Bueno, E., Mesa, S., Bedmar, E. J., Richardson, D. J., & Delgado, M. J. (2012).

Bacterial adaptation of respiration from oxic to microoxic and anoxic

conditions: redox control. Antioxidants & Redox Signaling, 16(8), 819–52.

doi:10.1089/ars.2011.4051

Centre, H., & Office, M. (2013). International symposium on the stabilisation of

greenhouse gases, (February), 1–16. Retrieved from

papers2://publication/uuid/F9A1E5A7-A145-4BBB-8A57-5E3BABF936D7

Chen, G., Kumar, A., Wyman, T. H., & Moran, C. P. (2006). Spo0A-dependent

activation of an extended -10 region promoter in Bacillus subtilis. Journal of

Bacteriology, 188(4), 1411–8. doi:10.1128/JB.188.4.1411-1418.2006

Chen, J., Shen, J., Solem, C., & Jensen, P. R. (2015). A New Type of YumC-Like

Ferredoxin (Flavodoxin) Reductase Is Involved in Ribonucleotide Reduction.

mBio, 6(6), e01132–15. doi:10.1128/mBio.01132-15

175

Chen, J., Zhang, Z., Zhang, C., & Yu, B. (2015). Genome sequence of Geobacillus

thermoglucosidasius DSM2542, a platform hosts for biotechnological

applications with industrial potential. Journal of Biotechnology, 216, 98–9.

doi:10.1016/j.jbiotec.2015.10.002

Cripps, R. E., Eley, K., Leak, D. J., Rudd, B., Taylor, M., Todd, M., … Atkinson,

T. (2009). Metabolic engineering of Geobacillus thermoglucosidasius for high

yield ethanol production, 11, 398–408. doi:10.1016/j.ymben.2009.08.005

Cruz Ramos, H., Boursier, L., Moszer, I., Kunst, F., Danchin, A., & Glaser, P.

(1995). Anaerobic transcription activation in Bacillus subtilis: identification of

distinct FNR-dependent and -independent regulatory mechanisms. The EMBO

Journal, 14(23), 5984–94. Retrieved from



Cruz Ramos, H., Hoffmann, T., Marino, M., Nedjari, H., Presecan-Siedel, E.,

Dreesen, O., … Jahn, D. (2000). Fermentative Metabolism of Bacillus subtilis:

Physiology and Regulation of Gene Expression. Journal of Bacteriology, 182(11),

3072–3080. doi:10.1128/JB.182.11.3072-3080.2000

Dammeyer, T., & Tinnefeld, P. (2012). Engineered fluorescent proteins illuminate

the bacterial periplasm. Computational and Structural Biotechnology Journal, 3,

e201210013. doi:10.5936/csbj.201210013

Dong, H., & Zhang, D. (2014). Current development in genetic engineering

strategies of Bacillus species. Microbial Cell Factories, 13, 63. doi:10.1186/1475-

2859-13-63

Durand, S., Braun, F., Lioliou, E., Romilly, C., Helfer, A.-C., Kuhn, L., …

Condon, C. (2015). A nitric oxide regulated small RNA controls expression of

genes involved in redox homeostasis in Bacillus subtilis. PLoS Genetics, 11(2),

e1004957. doi:10.1371/journal.pgen.1004957


176

Eijlander, R. T., Jongbloed, J. D. H., & Kuipers, O. P. (2009). Relaxed specificity

of the Bacillus subtilis TatAdCd translocase in Tat-dependent protein

secretion. Journal of Bacteriology, 191(1), 196–202. doi:10.1128/JB.01264-08

Esbelin, J., Armengaud, J., Zigha, A., & Duport, C. (2009). ResDE-dependent

regulation of enterotoxin gene expression in Bacillus cereus: evidence for

multiple modes of binding for ResD and interaction with Fnr. Journal of

Bacteriology, 191(13), 4419–26. doi:10.1128/JB.00321-09

Fabret, C., Ehrlich, S. D., & Noirot, P. (2002). A new mutation delivery system for

genome-scale approaches in Bacillus subtilis. Molecular Microbiology, 46(1), 25–

36. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12366828

Farr, G. A., Oussenko, I. A., & Bechhofer, D. H. (1999). Protection against 3’-to-5'

RNA decay in Bacillus subtilis. Journal of Bacteriology, 181(23), 7323–30.

Retrieved from



Fischer, A. B. (1975). Gentamicin as a bactericidal antibiotic in tissue culture.

Medical Microbiology and Immunology, 161(1), 23–39. doi:10.1007/BF02120767

Fisher, A. C., & DeLisa, M. P. (2008). Laboratory evolution of fast-folding green

fluorescent protein using secretory pathway quality control. PloS One, 3(6),

e2351. doi:10.1371/journal.pone.0002351

Fong, R., Hu, Z., Hutchinson, C. R., Huang, J., Cohen, S., & Kao, C. (2007).

Characterization of a large, stable, high-copy-number Streptomyces plasmid

that requires stability and transfer functions for heterologous polyketide

overproduction. Applied and Environmental Microbiology, 73(4), 1296–1307.

doi:10.1128/AEM.01888-06

Fry, B., Zhu, T., Domach, M. M., Koepsel, R. R., Phalakornkule, C., & Ataai, M.

M. (2000). Characterization of growth and acid formation in a Bacillus subtilis

pyruvate kinase mutant. Applied and Environmental Microbiology, 66(9), 4045–9.

Retrieved from

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=92257&tool=pm

centrez&rendertype=abstract

Geng, H., Zhu, Y., Mullen, K., Zuber, C. S., & Nakano, M. M. (2007).

Characterization of ResDE-dependent fnr transcription in Bacillus subtilis.

Journal of Bacteriology, 189(5), 1745–55. doi:10.1128/JB.01502-06

Geng, H., Zuber, P., & Nakano, M. M. (2007). Regulation of respiratory genes by

ResD-ResE signal transduction system in Bacillus subtilis. Methods in

Enzymology, 422, 448–64. doi:10.1016/S0076-6879(06)22023-8




177

Gilman, J., & Love, J. (2016). Synthetic promoter design for new microbial chassis.

Biochemical Society Transactions, 44(3), 731–7. doi:10.1042/BST20160042

Goosens, V. J., Monteferrante, C. G., & van Dijl, J. M. (2014). The Tat system of

Gram-positive bacteria. Biochimica et Biophysica Acta, 1843(8), 1698–706.

doi:10.1016/j.bbamcr.2013.10.008

Gougoulias, C., Clark, J. M., & Shaw, L. J. (2014). The role of soil microbes in the

global carbon cycle: tracking the below-ground microbial processing of plant-

derived carbon for manipulating carbon dynamics in agricultural systems.

Journal of the Science of Food and Agriculture, 94(12), 2362–71.

doi:10.1002/jsfa.6577

Green, J., & Paget, M. S. (2004). Bacterial redox sensors. Nature Reviews.

Microbiology, 2(12), 954–66. doi:10.1038/nrmicro1022

Gregory, M. A., Till, R., & Smith, M. C. M. (2003). Integration Site for

Streptomyces Phage BT1 and Development of Site-Specific Integrating

Vectors, 185(17), 5320–5323. doi:10.1128/JB.185.17.5320

Gust, B., Rourke, S. O., Bird, N., Kieser, T., & Chater, K. (2004). Recombineering

in Streptomyces coelicolor, 1–22.

Gyan, S., Shiohira, Y., Sato, I., Takeuchi, M., & Sato, T. (2006). Regulatory loop

between redox sensing of the NADH/NAD(+) ratio by Rex (YdiH) and

oxidation of NADH by NADH dehydrogenase Ndh in Bacillus subtilis.

Journal of Bacteriology, 188(20), 7062–71. doi:10.1128/JB.00601-06

Haas, L. O., Cregg, J. M., & Gleeson, M. A. (1990). Development of an integrative

DNA transformation system for the yeast Candida tropicalis. J. Bacteriol.,

172(8), 4571–4577. Retrieved from

http://jb.asm.org/content/172/8/4571.abstract?ijkey=cf9a47802b16cefc092f6

1c39b76b60f481c47eb&keytype2=tf_ipsecsha

Hamilton, T. L., Bryant, D. A., & Macalady, J. L. (2016). The role of biology in

planetary evolution: cyanobacterial primary production in low-oxygen

Proterozoic oceans. Environmental Microbiology, 18(2), 325–40.

doi:10.1111/1462-2920.13118

Härtig, E., & Jahn, D. (2012). Regulation of the Anaerobic Metabolism in Bacillus

subtilis, 61, 195–216. doi:10.1016/B978-0-12-394423-8.00005-6

Hill, J., Nelson, E., Tilman, D., Polasky, S., & Tiffany, D. (2006). Environmental,

economic, and energetic costs and benefits of biodiesel and ethanol biofuels.

http://jb.asm.org/content/172/8/4571.abstract?ijkey=cf9a47802b16cefc092f6

178

Proceedings of the National Academy of Sciences of the United States of America,

103(30), 11206–10. doi:10.1073/pnas.0604600103

IEA. (2015). Energy and Climate Change. World Energy Outlook Special Report, 1–

200. doi:10.1038/479267b

Ismaiel, A. A., Zhu, C. X., Colby, G. D., & Chen, J. S. (1993). Purification and

characterization of a primary-secondary alcohol dehydrogenase from two

strains of Clostridium beijerinckii. J. Bacteriol., 175(16), 5097–5105. Retrieved

from http://jb.asm.org/content/175/16/5097.long

Jasaitis, A., Borisov, V. B., Belevich, N. P., Morgan, J. E., Konstantinov, A. A., &

Verkhovsky, M. I. (2000). Electrogenic Reactions of Cytochrome bd †.

Biochemistry, 39(45), 13800–13809. doi:10.1021/bi001165n

John, R. P., Nampoothiri, K. M., & Pandey, A. (2007). Fermentative production of

lactic acid from biomass: an overview on process developments and future

perspectives. Applied Microbiology and Biotechnology, 74(3), 524–34.

doi:10.1007/s00253-006-0779-6

Juo, A. S. R. K. F. (2003). Tropical Soils : Properties and Management for Sustainable

Agriculture: Properties and Management for Sustainable Agriculture. Oxford

University Press, USA. Retrieved from

https://books.google.com/books?id=dVGgQcuoibAC&pgis=1

Kananaviciute, R., & Citavicius, D. (2015). Genetic engineering of Geobacillus

spp. Journal of Microbiological Methods, 111, 31–9.

doi:10.1016/j.mimet.2015.02.002

Kavscek, M., Strazar, M., Curk, T., Natter, K., & Petrovc, U. (2015). Yeast as a

cell factory: current state and perspectives. Microbial Cell Factories, 14(1), 94.

doi:10.1186/s12934-015-0281-x

Korkhin, Y., Kalb(Gilboa), A. J., Peretz, M., Bogin, O., Burstein, Y., & Frolow, F.

(1998). NADP-dependent bacterial alcohol dehydrogenases: crystal structure,

cofactor-binding and cofactor specificity of the ADHs of Clostridium

beijerinckii and Thermoanaerobacter brockii. Journal of Molecular Biology,

278(5), 967–981. doi:10.1006/jmbi.1998.1750

Kreuzer, P., Gärtner, D., Allmansberger, R., & Hillen, W. (1989). Identification

and sequence analysis of the Bacillus subtilis W23 xylR gene and xyl operator.

Journal of Bacteriology, 171(7), 3840–5. Retrieved from



Kumar, P., Barrett, D. M., Delwiche, M. J., & Stroeve, P. (2009). Methods for

Pretreatment of Lignocellulosic Biomass for Efficient Hydrolysis and Biofuel

http://jb.asm.org/content/175/16/5097.long


179

Production. Industrial & Engineering Chemistry Research, 48(8), 3713–3729.

doi:10.1021/ie801542g

Lazazzera, B. a, Beinert, H., Khoroshilova, N., Kennedy, M. C., & Kiley, P. J.

(1996). DNA binding and dimerization of the Fe-S-containing FNR protein

from Escherichia coli are regulated by oxygen. The Journal of Biological

Chemistry, 271(5), 2762–8. Retrieved from


Lee, C. A., & Grossman, A. D. (2007). Identification of the origin of transfer (oriT)

and DNA relaxase required for conjugation of the integrative and conjugative

element ICEBs1 of Bacillus subtilis. Journal of Bacteriology, 189(20), 7254–7261.

doi:10.1128/JB.00932-07

Liao, H. H., & Kanikula, A. M. (1990). Increased efficiency of transformation

ofBacillus stearothermophilus by a plasmid carrying a thermostable kanamycin

resistance marker. Current Microbiology, 21(5), 301–306.

doi:10.1007/BF02092095

Liao, H., McKenzie, T., & Hageman, R. (1986). Isolation of a thermostable

enzyme variant by cloning and selection in a thermophile. Proceedings of the

National Academy of Sciences of the United States of America, 83(3), 576–80.

Retrieved from



Liu, B., Kearns, D. B., & Bechhofer, D. H. (2016). Expression of multiple Bacillus

subtilis genes is controlled by decay of slrA mRNA from Rho-dependent 3???

ends. Nucleic Acids Research, 44(7), 3364–3372. doi:10.1093/nar/gkw069

Liu, X., & De Wulf, P. (2004). Probing the ArcA-P modulon of Escherichia coli by

whole genome transcriptional analysis and sequence recognition profiling. The

Journal of Biological Chemistry, 279(13), 12588–97. doi:10.1074/jbc.M313454200

Lopes, M. S. G. (2015). Engineering biological systems toward a sustainable

bioeconomy. Journal of Industrial Microbiology & Biotechnology, 42(6), 813–38.

doi:10.1007/s10295-015-1606-9

Loui, C., Chang, A. C., & Lu, S. (2009). Role of the ArcAB two-component system

in the resistance of Escherichia coli to reactive oxygen stress. BMC

Microbiology, 9, 183. doi:10.1186/1471-2180-9-183

Ludwig, H., Homuth, G., Schmalisch, M., Dyka, F. M., Hecker, M., & Stülke, J.

(2001). Transcription of glycolytic genes and operons in Bacillus subtilis:

evidence for the presence of multiple levels of control of the gapA operon.




180

Luscombe, N. M., Laskowski, R. a, & Thornton, J. M. (2001). Amino acid-base

interactions: a three-dimensional analysis of protein-DNA interactions at an

atomic level. Nucleic Acids Research, 29(13), 2860–2874.

doi:10.1093/nar/29.13.2860

Lynd, L. R., van Zyl, W. H., McBride, J. E., & Laser, M. (2005). Consolidated

bioprocessing of cellulosic biomass: an update. Current Opinion in Biotechnology,

16(5), 577–83. doi:10.1016/j.copbio.2005.08.009

Malpica, R., Franco, B., Rodriguez, C., Kwon, O., & Georgellis, D. (2004).

Identification of a quinone-sensitive redox switch in the ArcB sensor kinase.

Proceedings of the National Academy of Sciences of the United States of America,

101(36), 13318–23. doi:10.1073/pnas.0403064101

Maranhao, A., & Ellington, A. D. (2016). Evolving orthogonal suppressor tRNAs

to incorporate modified amino acids. ACS Synthetic Biology.

doi:10.1021/acssynbio.6b00145

Marino, M., Hoffmann, T., Schmid, R., Möbitz, H., & Jahn, D. (2000). Changes in

protein synthesis during the adaptation of Bacillus subtilis to anaerobic growth

conditions. Microbiology (Reading, England), 146 ( Pt 1(1), 97–105.

doi:10.1099/00221287-146-1-97

Marreiros, B. C., Calisto, F., Castro, P. J., Duarte, A. M., Sena, F. V, Silva, A.

F., … Pereira, M. M. (2016). Exploring membrane respiratory chains.

Biochimica et Biophysica Acta, 1857(8), 1039–67.

doi:10.1016/j.bbabio.2016.03.028

Mazoch, J., & Kucera, I. (2002). Control of gene expression by FNR-like proteins

in facultatively anaerobic bacteria. Folia Microbiologica, 47(2), 95–103.

Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12058404

Mclaughlin, K. J., Strain-damerell, C. M., Xie, K., Brekasis, D., Soares, A. S.,

Paget, M. S. B., & Kielkopf, C. L. (2010). Article Structural Basis for NADH /

NAD + Redox Sensing by a Rex Family Repressor. MOLCEL, 38(4), 563–575.

doi:10.1016/j.molcel.2010.05.006

McLaughlin, K. J., Strain-Damerell, C. M., Xie, K., Brekasis, D., Soares, A. S.,

Paget, M. S. B., & Kielkopf, C. L. (2010). Structural Basis for

NADH/NAD+ Redox Sensing by a Rex Family Repressor. Molecular Cell,

38(4), 563–575. doi:10.1016/j.molcel.2010.05.006


181

Mettert, E. L., & Kiley, P. J. (2005). ClpXP-dependent proteolysis of FNR upon

loss of its O2-sensing [4Fe-4S] cluster. Journal of Molecular Biology, 354(2), 220–

32. doi:10.1016/j.jmb.2005.09.066

Miethke, M., Schmidt, S., & Marahiel, M. A. (2008). The major facilitator

superfamily-type transporter YmfE and the multidrug-efflux activator Mta

mediate bacillibactin secretion in Bacillus subtilis. Journal of Bacteriology,

190(15), 5143–52. doi:10.1128/JB.00464-08

Mikulskis, A., Aristarkhov, A., & Lin, E. C. (1997). Regulation of expression of the

ethanol dehydrogenase gene (adhE) in Escherichia coli by catabolite repressor

activator protein Cra. Journal of Bacteriology, 179(22), 7129–34. Retrieved from



Miller, R. G., & Sorrell, S. R. (2014). The future of oil supply. Philosophical

Transactions. Series A, Mathematical, Physical, and Engineering Sciences, 372(2006),

20130179. doi:10.1098/rsta.2013.0179

Morris, R. L., & Schmidt, T. M. (2013). Shallow breathing: bacterial life at low

O(2). Nature Reviews. Microbiology, 11(3), 205–12. doi:10.1038/nrmicro2970

Myers, K. S., Yan, H., Ong, I. M., Chung, D., Liang, K., Tran, F., … Kiley, P. J.

(2013). Genome-scale analysis of escherichia coli FNR reveals complex

features of transcription factor binding. PLoS Genetics, 9(6), e1003565.

doi:10.1371/journal.pgen.1003565

Nah, H.-J., Woo, M.-W., Choi, S.-S., & Kim, E.-S. (2015). Precise cloning and

tandem integration of large polyketide biosynthetic gene cluster using

Streptomyces artificial chromosome system. Microbial Cell Factories, 14(1), 140.

doi:10.1186/s12934-015-0325-2

Nakamura, A., Sosa, A., Komori, H., Kita, A., & Miki, K. (2007). Crystal structure

of TTHA1657 (AT-rich DNA-binding protein; p25) from Thermus

thermophilus HB8 at 2.16 A resolution. Proteins, 66(3), 755–9.

doi:10.1002/prot.21222

Nakano, M. M., Dailly, Y. P., & Zuber, P. (1997). Characterization of anaerobic

fermentative growth of Bacillus subtilis : identification of fermentation end

products and genes required for growth . Characterization of Anaerobic

Fermentative Growth of Bacillus subtilis : Identification of Fermentation En.

Nieves, L. M., Panyon, L. A., & Wang, X. (2015). Engineering Sugar Utilization

and Microbial Tolerance toward Lignocellulose Conversion. Frontiers in

Bioengineering and Biotechnology, 3, 17. doi:10.3389/fbioe.2015.00017

Pagels, M., Fuchs, S., Pané-Farré, J., Kohler, C., Menschner, L., Hecker, M., …

Engelmann, S. (2010). Redox sensing by a Rex-family repressor is involved in


182

the regulation of anaerobic gene expression in Staphylococcus aureus.


Poggi, D., Oliveira de Giuseppe, P., & Picardeau, M. (2010). Antibiotic resistance

markers for genetic manipulations of Leptospira spp. Applied and Environmental

Microbiology, 76(14), 4882–5. doi:10.1128/AEM.00775-10

Pompeo, F., Foulquier, E., & Galinier, A. (2016). Impact of Serine/Threonine

Protein Kinases on the Regulation of Sporulation in Bacillus subtilis. Frontiers

in Microbiology, 7, 568. doi:10.3389/fmicb.2016.00568

Ramírez, S., Moreno, R., Zafra, O., Castán, P., Vallés, C., & Berenguer, J. (2000).

Two nitrate/nitrite transporters are encoded within the mobilizable plasmid for

nitrate respiration of Thermus thermophilus HB8. Journal of Bacteriology,

182(8), 2179–83. Retrieved from



Ravcheev, D. a, Li, X., Latif, H., Zengler, K., Leyn, S. a, Korostelev, Y. D., …

Rodionov, D. a. (2012). Transcriptional regulation of central carbon and

energy metabolism in bacteria by redox-responsive repressor Rex. Journal of


Reents, H., Münch, R., Dammeyer, T., Härtig, E., Mu, R., Jahn, D., & Ha, E.

(2006). The Fnr Regulon of Bacillus subtilis The Fnr Regulon of Bacillus

subtilis †, 188(3). doi:10.1128/JB.188.3.1103


183

Reeve, B., Martinez-Klimova, E., de Jonghe, J., Leak, D. J., & Ellis, T. (2016). The

Geobacillus Plasmid Set: A Modular Toolkit for Thermophile Engineering.

ACS Synthetic Biology. doi:10.1021/acssynbio.5b00298

Richardson, D. J. (2000). Bacterial respiration: a flexible process for a changing

environment. Microbiology (Reading, England), 146 ( Pt 3(3), 551–71.

doi:10.1099/00221287-146-3-551

Rodionov, D. A., Dubchak, I. L., Arkin, A. P., Alm, E. J., & Gelfand, M. S.

(2005). Dissimilatory metabolism of nitrogen oxides in bacteria: comparative

reconstruction of transcriptional networks. PLoS Computational Biology, 1(5),

e55. doi:10.1371/journal.pcbi.0010055

Roggiani, M., & Goulian, M. (2015). Oxygen-Dependent Cell-to-Cell Variability in

the Output of the Escherichia coli Tor Phosphorelay. Journal of Bacteriology,

197(12), 1976–1987. doi:10.1128/JB.00074-15

Romero, S., Merino, E., Bolívar, F., Gosset, G., & Martinez, A. (2007). Metabolic

engineering of Bacillus subtilis for ethanol production: lactate dehydrogenase

plays a key role in fermentative metabolism. Applied and Environmental


Rosenfeld, N., Elowitz, M. B., & Alon, U. (2002). Negative Autoregulation Speeds

the Response Times of Transcription Networks. Journal of Molecular Biology,

323(5), 785–793. doi:10.1016/S0022-2836(02)00994-4

Saha, B. C. (2003). Hemicellulose bioconversion. Journal of Industrial Microbiology &

Biotechnology, 30(5), 279–91. doi:10.1007/s10295-003-0049-x

Shaw, a J., Podkaminer, K. K., Desai, S. G., Bardsley, J. S., Rogers, S. R.,

Thorne, P. G.Lynd, L. R. (2008). Metabolic engineering of a thermophilic

bacterium to produce ethanol at high yield. Proceedings of the National Academy

of Sciences of the United States of America, 105(37), 13769–74.

doi:10.1073/pnas.0801266105

Shimada, T., Yamamoto, K., & Ishihama, A. (2011). Novel members of the Cra

regulon involved in carbon metabolism in Escherichia coli. Journal of


Sickmier, E. A., Brekasis, D., Paranawithana, S., Bonanno, J. B., Paget, M. S. B.,

Burley, S. K., & Kielkopf, C. L. (2005). X-Ray Structure of a Rex-Family

Repressor / NADH Complex Insights into the Mechanism of Redox Sensing,

13, 43–54. doi:10.1016/j.str.2004.10.012

184

Silva-Rocha, R., Martínez-García, E., Calles, B., Chavarría, M., Arce-Rodríguez,

A., de Las Heras, A.de Lorenzo, V. (2013). The Standard European Vector

Architecture (SEVA): a coherent platform for the analysis and deployment

of complex prokaryotic phenotypes. Nucleic Acids Research, 41(Database

issue), D666–75. doi:10.1093/nar/gks1119

Simmons, B. A., Loque, D., & Blanch, H. W. (2008). Next-generation biomass

feedstocks for biofuel production. Genome Biology, 9(12), 242. doi:10.1186/gb-

2008-9-12-242

Simon, R., Priefer, U., & Pühler, A. (1983). A BROAD HOST RANGE

MOBILIZATION SYSTEM FOR INVIVO GENETIC-ENGINEERING -

TRANSPOSON MUTAGENESIS IN GRAM-NEGATIVE BACTERIA.

BIO-TECHNOLOGY, 1(9). Retrieved from https://pub.uni-

bielefeld.de/publication/1659456

Sims, R., Taylor, M., Jack, S., & Mabee, W. (2008). From 1st to 2nd Generation

Bio Fuel Technologies: An overview of current industry and RD&D activities.

IEA Bioenergy, (November), 1–124.

Sørensen, H. P., & Mortensen, K. K. (2005). Soluble expression of recombinant

proteins in the cytoplasm of Escherichia coli. Microbial Cell Factories, 4(1), 1.

doi:10.1186/1475-2859-4-1

Spaans, S. K., Weusthuis, R. A., van der Oost, J., & Kengen, S. W. M. (2015).

NADPH-generating systems in bacteria and archaea. Frontiers in Microbiology,

6, 742. doi:10.3389/fmicb.2015.00742

Stanton, B. C., Nielsen, A. A. K., Tamsir, A., Clancy, K., Peterson, T., & Voigt, C.

A. (2014). Genomic mining of prokaryotic repressors for orthogonal logic

gates. Nature Chemical Biology, 10(2), 99–105. doi:10.1038/nchembio.1411

Strain-damerell, C. M. (n.d.). Functional analysis of Rex , a sensor of the NADH /

NAD + redox poise in Streptomyces coelicolor.

Suzuki, H., Murakami, A., & Yoshida, K. (2012). Counterselection system for

Geobacillus kaustophilus HTA426 through disruption of pyrF and pyrR.

Applied and Environmental Microbiology, 78(20), 7376–83.

doi:10.1128/AEM.01669-12

Suzuki, H., & Yoshida, K. (2012). Genetic transformation of Geobacillus

kaustophilus HTA426 by conjugative transfer of host-mimicking plasmids.

Journal of Microbiology and Biotechnology, 22(9), 1279–87. Retrieved from


Tang, Y. J., Sapra, R., Joyner, D., Hazen, T. C., Myers, S., Reichmuth, D., …

Keasling, J. D. (2009). Analysis of metabolic pathways and fluxes in a newly


185

discovered thermophilic and ethanol-tolerant Geobacillus strain. Biotechnology

and Bioengineering, 102(5), 1377–86. doi:10.1002/bit.22181

Taylor, M. P., Eley, K. L., Martin, S., Tuffin, M. I., Burton, S. G., & Cowan, D.

A. (2009). Thermophilic ethanologenesis: future prospects for second-

generation bioethanol production. Trends in Biotechnology, 27(7), 398–405.

doi:10.1016/j.tibtech.2009.03.006

Taylor, M. P., Esteban, C. D., & Leak, D. J. (2008). Development of a versatile

shuttle vector for gene expression in Geobacillus spp. Plasmid, 60(1), 45–52.

doi:10.1016/j.plasmid.2008.04.001

Tobisch, S., Zühlke, D., Bernhardt, J., Stülke, J., & Hecker, M. (1999). Role of

CcpA in regulation of the central pathways of carbon catabolism in Bacillus

subtilis. Journal of Bacteriology, 181(22), 6996–7004. Retrieved from


centrez&rendertype=abstract

Tolla, D. A., & Savageau, M. A. (2010). Regulation of aerobic-to-anaerobic

transitions by the FNR cycle in Escherichia coli. Journal of Molecular Biology,

397(4), 893–905. doi:10.1016/j.jmb.2010.02.015

Tominaga, Y., Ohshiro, T., & Suzuki, H. (2016). Conjugative plasmid transfer

from Escherichia coli is a versatile approach for genetic transformation of

thermophilic Bacillus and Geobacillus species. Extremophiles : Life under Extreme

Conditions, 20(3), 375–81. doi:10.1007/s00792-016-0819-9

Trchounian, A. (2004). Escherichia coli proton-translocating F0F1-ATP synthase

and its association with solute secondary transporters and/or enzymes of

anaerobic oxidation-reduction under fermentation. Biochemical and Biophysical

Research Communications, 315(4), 1051–7. doi:10.1016/j.bbrc.2004.02.005

Tripathi, S. A., Olson, D. G., Argyros, D. A., Miller, B. B., Barrett, T. F., Murphy,

D. M.,Caiazza, N. C. (2010). Development of pyrF-based genetic system for

targeted gene deletion in Clostridium thermocellum and creation of a pta

mutant. Applied and Environmental Microbiology, 76(19), 6591–9.

doi:10.1128/AEM.01484-10

Turner, R. J., Bonner, E. R., Grabner, G. K., & Switzer, R. L. (1998). Purification

and Characterization of Bacillus subtilis PyrR, a Bifunctional pyr mRNA-

binding Attenuation Protein/Uracil Phosphoribosyltransferase. Journal of

Biological Chemistry, 273(10), 5932–5938. doi:10.1074/jbc.273.10.5932


186

Tuteja, R. (2005). Type I signal peptidase: an overview. Archives of Biochemistry and

Biophysics, 441(2), 107–11. doi:10.1016/j.abb.2005.07.013

Unden, G., & Bongaerts, J. (1997). Alternative respiratory pathways of Escherichia

coli: energetics and transcriptional regulation in response to electron acceptors.

Biochimica et Biophysica Acta (BBA) - Bioenergetics, 1320(3), 217–234.

doi:10.1016/S0005-2728(97)00034-0

Unden, G., Steinmetz, P. A., & Degreif-Dünnwald, P. (2014). The Aerobic and

Anaerobic Respiratory Chain of Escherichia coli and Salmonella enterica:

Enzymes and Energetics. EcoSal Plus, 6(1). doi:10.1128/ecosalplus.ESP-0005-

2013

Volbeda, A., Darnault, C., Renoux, O., Nicolet, Y., & Fontecilla-Camps, J. C.

(2015). The crystal structure of the global anaerobic transcriptional regulator

FNR explains its extremely fine-tuned monomer-dimer equilibrium. Science

Advances, 1(11), e1501086. doi:10.1126/sciadv.1501086

Voskuil, M. I., Voepel, K., & Chambliss, G. H. (1995). The ? 16 region, a vital

sequence for the utilization of a promoter in Bacillus subtilis and Escherichia

coli. Molecular Microbiology, 17(2), 271–279. doi:10.1111/j.1365-

2958.1995.mmi_17020271.x

Wang, E., Bauer, M. C., Rogstam, A., Linse, S., Logan, D. T., & von

Wachenfeldt, C. (2008). Structure and functional properties of the Bacillus

subtilis transcriptional repressor Rex. Molecular Microbiology, 69(2), 466–78.

doi:10.1111/j.1365-2958.2008.06295.x

Wang, E., Ikonen, T. P., Knaapila, M., Svergun, D., Logan, D. T., & Wachenfeldt,

C. Von. (2011). Small-angle X-ray Scattering Study of a Rex Family

Repressor : Conformational Response to NADH and NAD + Binding in

Solution. Journal of Molecular Biology, 408(4), 670–683.

doi:10.1016/j.jmb.2011.02.050

Weber, J., & Senior, A. E. (2003). ATP synthesis driven by proton transport in F 1

F 0 -ATP synthase. FEBS Letters, 545(1), 61–70. doi:10.1016/S0014-

5793(03)00394-6

187

WEF. (2016). The global risks report 2016, 11th edition, 103. Retrieved from

https://www.weforum.org/reports/the-global-risks-report-2016/

Weiner, J. H., Bilous, P. T., Shaw, G. M., Lubitz, S. P., Frost, L., Thomas, G.

H., … Turner, R. J. (1998). A novel and ubiquitous system for membrane

targeting and secretion of cofactor-containing proteins. Cell, 93(1), 93–101.

doi:10.1016/S0092-8674(00)81149-6

Wietzke, M., & Bahl, H. (2012). The redox-sensing protein Rex, a transcriptional

regulator of solventogenesis in Clostridium acetobutylicum. Applied

Microbiology and Biotechnology, 96(3), 749–61. doi:10.1007/s00253-012-4112-2

Williams, P. R. D., Inman, D., Aden, A., & Heath, G. A. (2009). Environmental

and sustainability factors associated with next-generation biofuels in the U.S.:

what do we really know? Environmental Science & Technology, 43(13), 4763–75.

Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/19673263

Xu, P., Li, L., Zhang, F., Stephanopoulos, G., & Koffas, M. (2014). Improving

fatty acids production by engineering dynamic pathway regulation and

metabolic control. Proceedings of the National Academy of Sciences of the United

States of America, 111(31), 11299–304. doi:10.1073/pnas.1406401111

Yagi, T. (1993). The bacterial energy-transducing NADH-quinone oxidoreductases.

Biochimica et Biophysica Acta (BBA) - Bioenergetics, 1141(1), 1–17.

doi:10.1016/0005-2728(93)90182-F

Yamamoto, S., Gunji, W., Suzuki, H., Toda, H., Suda, M., Jojima, T., …

Yukawa, H. (2012). Overexpression of genes encoding glycolytic enzymes in

Corynebacterium glutamicum enhances glucose metabolism and alanine

production under oxygen deprivation conditions. Applied and Environmental


Yasueda, H., Kawahara, Y., & Sugimoto, S. (1999). Bacillus subtilis yckG and

yckF Encode Two Key Enzymes of the Ribulose Monophosphate Pathway

Used by Methylotrophs, and yckH Is Required for Their Expression. J.

Bacteriol., 181(23), 7154–7160. Retrieved from

http://jb.asm.org/content/181/23/7154?ijkey=ea11518321460e880a5d6a0e8

45157ef919f044f&keytype2=tf_ipsecsha

Yu, J., & Tao, M. (2010). [Effect of inorganic salts on the conjugation and

heterologous expression of actinorhodin in Streptomyces avermitilis]. Wei

Sheng Wu Xue Bao = Acta Microbiologica Sinica, 50(11), 1556–61. Retrieved from


http://www.weforum.org/reports/the-global-risks-report-2016/


http://jb.asm.org/content/181/23/7154?ijkey=ea11518321460e880a5d6a0e8


188

Yurimoto, H., Hirai, R., Matsuno, N., Yasueda, H., Kato, N., & Sakai, Y. (2005).

HxlR, a member of the DUF24 protein family, is a DNA-binding protein that

acts as a positive regulator of the formaldehyde-inducible hxlAB operon in

Bacillus subtilis. Molecular Microbiology, 57(2), 511–9. doi:10.1111/j.1365-

2958.2005.04702.x

Zeigler, D. R. (2014). The Geobacillus paradox: why is a thermophilic bacterial

genus so prevalent on a mesophilic planet? Microbiology (Reading, England),

160(Pt 1), 1–11. doi:10.1099/mic.0.071696-0

Zeldes, B. M., Keller, M. W., Loder, A. J., Straub, C. T., Adams, M. W. W., &

Kelly, R. M. (2015). Extremely thermophilic microorganisms as metabolic

engineering platforms for production of fuels and industrial chemicals. Frontiers

in Microbiology, 6, 1209. doi:10.3389/fmicb.2015.01209

Zhang, L., Nie, X., Ravcheev, D. a, Rodionov, D. a, Sheng, J., Gu, Y., … Yang,

C. (2014). Redox-responsive repressor Rex modulates alcohol production and

oxidative stress tolerance in Clostridium acetobutylicum. Journal of Bacteriology,

196(22), 3949–63. doi:10.1128/JB.02037-14

Zhang, X., Shanmugam, K. T., & Ingram, L. O. (2010). Fermentation of glycerol

to succinate by metabolically engineered strains of Escherichia coli. Applied and

Environmental Microbiology, 76(8), 2397–401. doi:10.1128/AEM.02902-09

Zheng, Y., Ko, T. P., Sun, H., Huang, C. H., Pei, J., Qiu, R.,Guo, R. T. (2014).

Distinct structural features of Rex-family repressors to sense redox levels in

anaerobes and aerobes. Journal of Structural Biology, 188(3), 195–204.

doi:10.1016/j.jsb.2014.11.001

189

8. Appendix

Abbreviations

AprR Apramycin resistant

ATP Adenosine 5’-triphosphate

BLAST Basic local alignment tool

Bp Base pair

B-Rex Bacillus subtilis Rex

ChIP Chromatin immunoprecipitation

C-Rex Clostridium thermocellum Rex

Da Daltons

dH2O distilled water

DNA Deoxyribonucleic acid

dNTP Deoxyribonucleoside triphosphate

EMSA Electromobility shift assay

G-Rex Geobacillus thermoglucosidans Rex

g Grams

h Hour(s)

KanR Kanamycin resistant

KanS Kanamycin sensitive

L Litre

min Minute(s)

NAD+ Nicotinamide adenine dinucleotide, oxidised form

NADH Nicotinamide adenine dinucleotide, reduced form

OD Optical density (wavelength indicated in subscript)

190

PCR Polymerase chain reaction

PDB Protein data bank

PMF Proton motive force

qPCR Quantitative PCR

ROP Rex operator

RT-qPCR Reverse transcriptase qPCR

sec Second(s)

T-Rex Thermus aquaticus Rex

U Unit(s)

DNA sequences

C-Rex 1 ATGAACCTGG ACAAAAAAAT CAGCATGGCA GTTATTCGTC GTCTGCCTCG TTATTATCGT

61 TATCTGAGCG ATCTGCTGAA ACTGGGTATT ACCCGTATTA GCAGCAAAGA ACTGAGCAGC

121 CGTATGGGCA TTACCGCAAG CCAGATTCGT CAGGATCTGA ATTGTTTTGG TGGTTTTGGT

181 CAGCAGGGTT ATGGTTATAA TGTGGAATAT CTGTACAAAG AGATCGGCAA TATTCTGGGT

241 GTTAATGAGG CCTTCAAAAT CATTATTATC GGTGCCGGTA ATATGGGTCA GGCACTGGCA

301 AATTATACCA ACTTTGAAAA ACGCGGTTTC AAACTGATCG GCATCTTTGA CATTAATCCG

361 AACCTGATCG GGAAAAAAAT CCGTGATGTT GAAATCATGC ATCTGGATAG CCTGGATCGT

421 TTTGTTGCAG AAAATCAGGT TGATATTGCC ATTCTGTGTG TGCCGTATGA AAATACACCG

481 GCAGTTGCAG ATAAAGTTGC ACGTCTGGGT GTGAAAGGTC TGTGGAATTT TAGCCCGATG

541 GATCTGAAAC TGCCTTATGA TGTGATTATC GAAAATGTGC ATCTGTCCGA TAGCCTGATG

601 GTTCTGGGTT ATCGTCTGAA TGAAATGCGT AAAAGCCAGC GCAATAAAAG CGACTACAAA

661 GATCATGACG GCGATTATAA GGACCACGAT ATCGACTACA AAGACGATGA CGATAAGTAA

721 c

G-Rex

1 ATGAACAATG AGCAGCCAAA GATCCCACAG GCAACTGCTA AACGTCTGCC TCTGTATTAT

61 CGTTTTCTGA AGAATTTACA TGCTTCAGGA AAACAGCGTG TTAGCAGCGC GGAATTATCT

121 GAAGCCGTCA AAGTAGATAG CGCAACAATT CGCCGCGACT TCTCATATTT TGGCGCTCTG

181 GGTAAAAAGG GCTATGGCTA CAACGTTAAC TATCTGCTGT CTTTTTTCCG GAAAACACTG

241 GATCAGGACG AAATCACGGA AGTTGCTCTT TTCGGCGTTG GCAACCTTGG TACGGCATTC

301 TTAAATTACA ATTTCAGCAA AAATAACAAT ACAAAAATCG TAATTGCGTT TGACGTGGAT

361 GAGGAGAAAA TTGGTAAAGA GGTTGGTGGG GTGCCTGTAT ATCACCTTGA CGAAATGGAA

421 ACAAGACTCC ACGAAGGCAT TCCAGTTGCC ATTTTAACAG TTCCGGCACA TGTCGCGCAA

481 TGGATTACTG ACCGACTTGT TCAAAAAGGC ATCAAAGGCA TTTTGAACTT CACACCGGCA

541 AGACTTAATG TCCCGAAACA TATTAGAGTT CATCATATCG ACCTGGCAGT CGAATTGCAG

601 TCGTTAGTTT ATTTTTTGAA AAACTATCCA GCGGAGGACT ACAAAGATCA TGACGGCGAT

661 TATAAGGACC ACGATATCGA CTACAAAGAC GATGACGATA AGTAAc

Documents

of Sussex PhD thesissro.sussex.ac.uk/67293/1/Macklyne, Heather-Rose Victoria.pdf · enable strain development for second generation biofuel production. We have focused on the redox-sensing